From matthias.andree at gmx.de  Thu Oct  1 11:51:50 2009
From: matthias.andree at gmx.de (Matthias Andree)
Date: Thu, 01 Oct 2009 12:51:50 +0200
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <4AC2692B.8090703@thekelleys.org.uk>
References: <4AC2692B.8090703@thekelleys.org.uk>
Message-ID: <op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>

Am 29.09.2009, 22:08 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:

> The development branch of dnsmasq has amassed quite a few changes, going
> back 3 months in some cases: most of the work was done before the 2.50
> security-only release. I think it's time therefore to make a new
> release, and I'm making a first release-candidate available at
>
> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.gz
> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.lzma
>
> Please test this, and/or translate it, as appropriate. (Translators,
> drop me a note if you're working and I'll hold off a final release until
> you're done.)
>
> version 2.51
>             Add support for internationalised DNS. Non-ASCII characters
>             in domain names found in /etc/hosts, /etc/ethers and
>             /etc/dnsmasq.conf will be correctly handled by translation
>             to punycode, as specified in RFC3490. This function is only
>             available if dnsmasq is compiled with internationalisation
>             support, and adds a dependency on GNU libidn. Without i18n
>             support, dnsmasq continues to be compilable with just
>             standard tools. Thanks to Yves Dorfsman for the
>             suggestion.

Some findings of the -rc1:

- During build, I18N - which currently comprises IDN _and_ NLS as an  
inseparable union - only works with GNU make, not with BSD's make.

- IDN needs pkg-config (just like DBus also needs pkg-config) - there are  
some GNUisms, but I don't have good/efficiently implemented ideas to avoid  
these. GNU make is readily available on most systems though.

These two requirements are not documented, please consider documenting  
them.

- Unless I'm missing some links in setup.html, the whole build is  
massively underdocumented, no mention of DBus or i18n anywhere, and  
setup.html doesn't appear to be linked from doc.html and vice versa. I'd  
suggest to add the whole COPT=-Dfoo system to setup.html, too, or at least  
point to src/config.h and mention how to most efficiently set these  
options.

- I wonder how dnsmasq will figure out the proper character set for these  
files -- or will it just assume UTF-8? Older systems might use some  
ISO-8859-* variant. (Sorry, no time to read source now.)

- The "all-i18n" target creates the .mo files under /src which I find  
counterintuitive.

- The German translation is - with 17/313 messages - so incomplete it's  
not worth shipping. I don't have sufficient time to do it on short notice  
though, so I'll not try.

Best regards
Matthias

-- 
Matthias Andree


From matthias.andree at gmx.de  Thu Oct  1 11:57:45 2009
From: matthias.andree at gmx.de (Matthias Andree)
Date: Thu, 01 Oct 2009 12:57:45 +0200
Subject: [Dnsmasq-discuss] dnsmasq 2.51 rc1 for FreeBSD (was: Announce:
 dnsmasq-2.51rc1)
In-Reply-To: <4AC2692B.8090703@thekelleys.org.uk>
References: <4AC2692B.8090703@thekelleys.org.uk>
Message-ID: <op.u04bejyf1e62zd@balu.cs.uni-paderborn.de>

Am 29.09.2009, 22:08 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:

> The development branch of dnsmasq has amassed quite a few changes, going
> back 3 months in some cases: most of the work was done before the 2.50
> security-only release. I think it's time therefore to make a new
> release, and I'm making a first release-candidate available at
>
> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.gz
> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.lzma

Greetings,

for FreeBSD, I have submitted a 2.51rc1 port update for ports/dns/dnsmasq,  
which will enable IDN and NLS by default and pull in gettext-related  
packages, iconv, libidn, and gmake. You can change that in the "make  
config" dialogue.

At time of this writing, the update has not yet been committed to the  
ports tree; in the meanwhile you can get the patch against the 2.50  
version in the ports tree from the problem report tracker at  
<http://www.freebsd.org/cgi/query-pr.cgi?pr=139273> (deep link to the  
patch:  
<http://www.freebsd.org/cgi/query-pr.cgi?prp=139273-1-txt&n=/dnsmasq-2.51.r1.patch>).

Please send me feedback that are port-specific off-list (or at least Cc:  
me), and general feedback on the software to Simon or the list.

Thank you.

Best regards

-- 
Matthias Andree


From simon at thekelleys.org.uk  Thu Oct  1 13:19:02 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 01 Oct 2009 13:19:02 +0100
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
References: <4AC2692B.8090703@thekelleys.org.uk>
	<op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
Message-ID: <4AC49E36.6030507@thekelleys.org.uk>

Matthias Andree wrote:
> Am 29.09.2009, 22:08 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:
> 
>> The development branch of dnsmasq has amassed quite a few changes, going
>> back 3 months in some cases: most of the work was done before the 2.50
>> security-only release. I think it's time therefore to make a new
>> release, and I'm making a first release-candidate available at
>>
>> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.gz 
>>
>> http://www.thekelleys.org.uk/dnsmasq/release-candidates/dnsmasq-2.51rc1.tar.lzma 
>>
>>
>> Please test this, and/or translate it, as appropriate. (Translators,
>> drop me a note if you're working and I'll hold off a final release until
>> you're done.)
>>
>> version 2.51
>>             Add support for internationalised DNS. Non-ASCII characters
>>             in domain names found in /etc/hosts, /etc/ethers and
>>             /etc/dnsmasq.conf will be correctly handled by translation
>>             to punycode, as specified in RFC3490. This function is only
>>             available if dnsmasq is compiled with internationalisation
>>             support, and adds a dependency on GNU libidn. Without i18n
>>             support, dnsmasq continues to be compilable with just
>>             standard tools. Thanks to Yves Dorfsman for the
>>             suggestion.
> 
> Some findings of the -rc1:
> 
> - During build, I18N - which currently comprises IDN _and_ NLS as an 
> inseparable union - only works with GNU make, not with BSD's make.

Where is the problem? I ensured that the build works with BSD make, but 
only for the non-IDN build, it would be good to fix that.

> 
> - IDN needs pkg-config (just like DBus also needs pkg-config) - there 
> are some GNUisms, but I don't have good/efficiently implemented ideas to 
> avoid these. GNU make is readily available on most systems though.
> 
> These two requirements are not documented, please consider documenting 
> them.
> 
> - Unless I'm missing some links in setup.html, the whole build is 
> massively underdocumented, no mention of DBus or i18n anywhere, and 
> setup.html doesn't appear to be linked from doc.html and vice versa. I'd 
> suggest to add the whole COPT=-Dfoo system to setup.html, too, or at 
> least point to src/config.h and mention how to most efficiently set 
> these options.
> 

A valid criticism, I guess that since almost everyone installs dnsmasq 
from an OS or distribution package these days, such documenation has a 
much smaller audience (and therefore priority) than user-level 
documentation. It is sadly lacking.



> - I wonder how dnsmasq will figure out the proper character set for 
> these files -- or will it just assume UTF-8? Older systems might use 
> some ISO-8859-* variant. (Sorry, no time to read source now.)

It calls idna_to_ascii_lz() which uses the current locale.

from "man idna_to_ascii_lz"

DESCRIPTION
        Convert  domain  name  in  the locale?s encoding to ASCII 
string.
        The domain name may contain several labels, separated by
        dots.  The output buffer must be deallocated by the caller.

> 
> - The "all-i18n" target creates the .mo files under /src which I find 
> counterintuitive.

I can't remember why it's done like that. It must have made sense at the 
time!

> 
> - The German translation is - with 17/313 messages - so incomplete it's 
> not worth shipping. I don't have sufficient time to do it on short 
> notice though, so I'll not try.

Any volunteers on the list?

Cheers,

Simon.


From matthias.andree at gmx.de  Thu Oct  1 15:00:56 2009
From: matthias.andree at gmx.de (Matthias Andree)
Date: Thu, 01 Oct 2009 16:00:56 +0200
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <4AC49E36.6030507@thekelleys.org.uk>
References: <4AC2692B.8090703@thekelleys.org.uk>
	<op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
	<4AC49E36.6030507@thekelleys.org.uk>
Message-ID: <op.u04jvuza1e62zd@balu.cs.uni-paderborn.de>

Am 01.10.2009, 14:19 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:

>>  - During build, I18N - which currently comprises IDN _and_ NLS as an  
>> inseparable union - only works with GNU make, not with BSD's make.
>
> Where is the problem? I ensured that the build works with BSD make, but  
> only for the non-IDN build, it would be good to fix that.

Dear Simon,

The problem lies in bld/Makefile which uses %-pattern rules (a GNUism) on  
line 16 rather than traditional suffix rules (lines 7 - 8).

      1  CFLAGS = -Wall -W -O2
      2
      3  OBJS = cache.o rfc1035.o util.o option.o forward.o network.o \
      4         dnsmasq.o dhcp.o lease.o rfc2131.o netlink.o dbus.o bpf.o \
      5         helper.o tftp.o log.o
      6
      7  .c.o:
      8          $(CC) $(CFLAGS) $(COPTS) $(I18N) $(DNSMASQ_CFLAGS)  
$(RPM_OPT_FLAGS) -c $<
      9
     10  dnsmasq : $(OBJS)
     11          $(CC) $(LDFLAGS) -o $@  $(OBJS) $(DNSMASQ_LIBS) $(LIBS)
     12
     13  dnsmasq.pot : $(OBJS:.o=.c) dnsmasq.h config.h
     14          $(XGETTEXT) -d dnsmasq --foreign-user --omit-header  
--keyword=_ -o $@ -i $(OBJS:.o=.c)
     15
     16  %.mo : ../po/%.po dnsmasq.pot
     17          $(MSGMERGE) -o - ../po/$*.po dnsmasq.pot | $(MSGFMT) -o  
$*.mo -

ll 13f: The ":.o=.c" modifier is supported on SysV and BSD, but I'm not  
sure about its portability to other systems.

l 16: The % patterns are unsupported in BSD make; in a traditional .po.mo:  
pattern you cannot directly list the prerequisite dnsmasq.pot, and if you  
list a $(MAKE) dnsmasq.pot you'll break parallel builds (make -j3) (so  
better list it in the command block of another rule instead), and I'm not  
sure how you could express the ../po/ path insert at all. Perhaps a  
POFILES=da.po de.po es pt_BR ro... style variable which you substitue from  
might help a bit, too.

Relevant documentation:
- Tutorial http://www.freebsd.org/doc/en_US.ISO8859-1/books/pmake/
- Manpage   
http://www.freebsd.org/cgi/man.cgi?query=make&apropos=0&sektion=0&manpath=FreeBSD+6.3-RELEASE&format=html

Note that overriding CFLAGS (line 1) is frowned upon in BSD, too, although  
I think DNSMASQ_CFLAGS and RPM_OPT_FLAGS offers plenty ways to merge  
things.

[IDN]
>> - I wonder how dnsmasq will figure out the proper character set for  
>> these files -- or will it just assume UTF-8? Older systems might use  
>> some ISO-8859-* variant. (Sorry, no time to read source now.)
>
> It calls idna_to_ascii_lz() which uses the current locale.

...which derives from implicit assumptions that the user starting dnsmasq  
uses the proper locale for the /etc/ files subjected to IDN/Punycode  
handling, particularly on systems where root users don't usually log in,  
but use sudo (Ubuntu Linux) - the console locale used to edit files may  
differ from the locale in effect when /etc/init.d/dnsmasq or init or  
Upstart or whichever launches the daemon.

This should be documented very clearly in the manpage and the relevant  
.html file to avoid surprises - the locale used to edit these /etc/ files  
must get set in the environment where dnsmasq is launched.

>>  - The "all-i18n" target creates the .mo files under /src which I find  
>> counterintuitive.
>
> I can't remember why it's done like that. It must have made sense at the  
> time!

Well at least they are side to side with the .o files which also end up in  
src/, and then I wonder what the bld/ directory is for if it's static. I  
would've guessed from the name that bld/ is build/ and hence where the  
outputs are supposed to end up, but apparently not so. :)

Cheers

-- 
Matthias Andree


From RLynn at fundamentalsltd.co.uk  Fri Oct  2 11:05:03 2009
From: RLynn at fundamentalsltd.co.uk (Roger Lynn)
Date: Fri, 02 Oct 2009 11:05:03 +0100
Subject: [Dnsmasq-discuss] FAQ entry - Debian hostnames
Message-ID: <4AC5D04F.4040703@fundamentalsltd.co.uk>

Hi,

In the dnsmasq FAQ there's a question: "I'm running Debian, and my machines get
an address fine with DHCP, but their names are not appearing in the DNS." The
answer includes "Update: the lastest dhcpcd packages _do_ send the hostname by
default." This isn't true for Debian, unless you are talking about the
experimental packages of ISC DHCP 4.1.

A patch has apparently been applied in Ubuntu, but the Debian maintainers are
waiting for upstream to fix it. See http://bugs.debian.org/151820 and
http://bugs.debian.org/522889 for details.

Thanks for your work on dnsmasq,

Roger



From simon at thekelleys.org.uk  Fri Oct  2 13:19:50 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Fri, 02 Oct 2009 13:19:50 +0100
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <op.u04jvuza1e62zd@balu.cs.uni-paderborn.de>
References: <4AC2692B.8090703@thekelleys.org.uk>
	<op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
	<4AC49E36.6030507@thekelleys.org.uk>
	<op.u04jvuza1e62zd@balu.cs.uni-paderborn.de>
Message-ID: <4AC5EFE6.7070202@thekelleys.org.uk>

Matthias Andree wrote:

> [IDN]
>>> - I wonder how dnsmasq will figure out the proper character set for 
>>> these files -- or will it just assume UTF-8? Older systems might use 
>>> some ISO-8859-* variant. (Sorry, no time to read source now.)
>>
>> It calls idna_to_ascii_lz() which uses the current locale.
> 
> ...which derives from implicit assumptions that the user starting 
> dnsmasq uses the proper locale for the /etc/ files subjected to 
> IDN/Punycode handling, particularly on systems where root users don't 
> usually log in, but use sudo (Ubuntu Linux) - the console locale used to 
> edit files may differ from the locale in effect when /etc/init.d/dnsmasq 
> or init or Upstart or whichever launches the daemon.
> 
> This should be documented very clearly in the manpage and the relevant 
> .html file to avoid surprises - the locale used to edit these /etc/ 
> files must get set in the environment where dnsmasq is launched.

I had assumed that the system locale would be set by and inherited from 
init, but looking more closely, that's not true. On Debian (and Ubuntu) 
the system default locale is stored in /etc/default/locale, and the 
startup scripts for daemons which care about this have to source that 
file and export LANG. I've fixed dnsmasq's Debian init script to do 
this, and that will feed through to Ubuntu too. Other packagers need to 
take this into account. I'll add to the documentation.

> 
>>>  - The "all-i18n" target creates the .mo files under /src which I 
>>> find counterintuitive.
>>
>> I can't remember why it's done like that. It must have made sense at 
>> the time!
> 
> Well at least they are side to side with the .o files which also end up 
> in src/, and then I wonder what the bld/ directory is for if it's 
> static. I would've guessed from the name that bld/ is build/ and hence 
> where the outputs are supposed to end up, but apparently not so. :)
> 

/bld is a place to keep build infrastructure, rather than cluttering up 
the root directory.


Cheers,

Simon.





From matthias.andree at gmx.de  Fri Oct  2 13:35:27 2009
From: matthias.andree at gmx.de (Matthias Andree)
Date: Fri, 02 Oct 2009 14:35:27 +0200
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <4AC5EFE6.7070202@thekelleys.org.uk>
References: <4AC2692B.8090703@thekelleys.org.uk>
	<op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
	<4AC49E36.6030507@thekelleys.org.uk>
	<op.u04jvuza1e62zd@balu.cs.uni-paderborn.de>
	<4AC5EFE6.7070202@thekelleys.org.uk>
Message-ID: <op.u06aldas1e62zd@balu.cs.uni-paderborn.de>

Dear Simon, list subscribers,

Am 02.10.2009, 14:19 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:

> I had assumed that the system locale would be set by and inherited from  
> init, but looking more closely, that's not true. On Debian (and Ubuntu)  
> the system default locale is stored in /etc/default/locale, and the  
> startup scripts for daemons which care about this have to source that  
> file and export LANG. I've fixed dnsmasq's Debian init script to do  
> this, and that will feed through to Ubuntu too. Other packagers need to  
> take this into account. I'll add to the documentation.

Thank you.

I wonder - also from experiences drawn from my own projects such as  
fetchmail - how much of documentation, and for which target group in  
respect, is required.
It appears to me that decent projects will benefit from a documentation  
that addresses end users as well as packagers.

One more train of thought (a bit dialectic), if you read a system-wide  
locale configuration, that may still not match a user-specific locale  
setting; however this would likely be a less common configuration for  
dnsmasq. Also, given that many distributions are converging on UTF-8  
encodings, at least the character representation will be the same, so that  
documentation would likely suffice for dnsmasq.

> /bld is a place to keep build infrastructure, rather than cluttering up  
> the root directory.

Fine.

Cheers

-- 
Matthias Andree


From simon at thekelleys.org.uk  Fri Oct  2 13:50:30 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Fri, 02 Oct 2009 13:50:30 +0100
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51rc1
In-Reply-To: <op.u06aldas1e62zd@balu.cs.uni-paderborn.de>
References: <4AC2692B.8090703@thekelleys.org.uk>
	<op.u04a4ods1e62zd@balu.cs.uni-paderborn.de>
	<4AC49E36.6030507@thekelleys.org.uk>
	<op.u04jvuza1e62zd@balu.cs.uni-paderborn.de>
	<4AC5EFE6.7070202@thekelleys.org.uk>
	<op.u06aldas1e62zd@balu.cs.uni-paderborn.de>
Message-ID: <4AC5F716.8010101@thekelleys.org.uk>

Matthias Andree wrote:
> Dear Simon, list subscribers,
> 
> Am 02.10.2009, 14:19 Uhr, schrieb Simon Kelley <simon at thekelleys.org.uk>:
> 
>> I had assumed that the system locale would be set by and inherited 
>> from init, but looking more closely, that's not true. On Debian (and 
>> Ubuntu) the system default locale is stored in /etc/default/locale, 
>> and the startup scripts for daemons which care about this have to 
>> source that file and export LANG. I've fixed dnsmasq's Debian init 
>> script to do this, and that will feed through to Ubuntu too. Other 
>> packagers need to take this into account. I'll add to the documentation.
> 
> Thank you.
> 
> I wonder - also from experiences drawn from my own projects such as 
> fetchmail - how much of documentation, and for which target group in 
> respect, is required.
> It appears to me that decent projects will benefit from a documentation 
> that addresses end users as well as packagers.

There can never be too much documentation, and for dnsmasq most of the 
required documentation is needed by both packagers and users. The things 
that you identified as missing to be with build targets and variables, 
are probably things that users don't need, if they are using packaged 
software.
> 
> One more train of thought (a bit dialectic), if you read a system-wide 
> locale configuration, that may still not match a user-specific locale 
> setting; however this would likely be a less common configuration for 
> dnsmasq. Also, given that many distributions are converging on UTF-8 
> encodings, at least the character representation will be the same, so 
> that documentation would likely suffice for dnsmasq.
> 

One of the files in question is /etc/hosts. The format of that is well 
defined, so there's no way to add a charset tag to it. Since it's a 
system-wide configuration file, then the only valid assumption is that 
it uses the system-default locale, I think. When I was doing this work, 
I experimented with non-ascii characters in /etc/hosts. At least with 
gnu libc, the resolver library works fine, but I suspect it may not if 
the locale associated with a process has a different character encoding 
to that used in /etc/hosts.


Cheers,

Simon.


From hashimoh at tmtv.ne.jp  Sun Oct  4 04:05:25 2009
From: hashimoh at tmtv.ne.jp (Hisashi Hashimoto)
Date: Sun, 04 Oct 2009 12:05:25 +0900
Subject: [Dnsmasq-discuss] dnsmasq as DHCP server on vmware client machine
Message-ID: <4AC810F5.4000102@tmtv.ne.jp>

Hi,

  I am using Dnsmasq 2.45, a little bit old on my Local distribution.
I am trying to use dnsmasq as a DHCP server on vmware host machine
instead of using vmware original DHCP server, vmnet-dhcpd.
Why I prefer using dnsmasq is that vmnet-dhcpd is based on ISC DHCPD
2.0(old) and I prefer to use dnsmasq configuration.

 I have one problem now.
This proble appeared after I moved VMware workstation 6.5 from 5.0.
Problem is:
Dnsmasq does not look at the vmnet1, 4, 8(Virtual network on host
machine created by VMware).

  Why I think so, is :
(1) When I use vmnet-dhcpd or ISC DHCPD 3.0, they answered the query
from guest machines. but when I use dnsmasq, it did not answer.
   So, network interfaces had been created and work without problems.
(2) While I run ISC DHCPD, I started dnsmasq with DHCP-RANGE option,
    Dnsmasq did not make any claim that port had been used by ISC DHCPD.

Any suggestions are welcome.
---------------------------------------------------
Here is my configuration related DHCP.
interface=vmnet1
interface=vmnet4
interface=vmnet8
interface=lo
bind-interfaces
dhcp-range=192.168.219.101,192.168.219.128,255.255.255.0,12h
dhcp-range=192.168.140.101,192.168.140.254,255.255.255.0,12h
dhcp-range=192.168.167.101,192.168.167.254,255.255.255.0,12h

Below is output of ifconfig.
vmnet1    Link encap:Ethernet  HWaddr 00:50:56:C0:00:01
          inet addr:192.168.140.1  Bcast:192.168.140.255
          Mask:255.255.255.0
          inet6 addr: fe80::250:56ff:fec0:1/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)

vmnet4    Link encap:Ethernet  HWaddr 00:50:56:C0:00:04
          inet addr:192.168.219.1  Bcast:192.168.219.255
          Mask:255.255.255.0
          inet6 addr: fe80::250:56ff:fec0:4/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:25 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)

vmnet8    Link encap:Ethernet  HWaddr 00:50:56:C0:00:08
          inet addr:192.168.167.1  Bcast:192.168.167.255
          Mask:255.255.255.0
          inet6 addr: fe80::250:56ff:fec0:8/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)

Hashimoto Hisashi


From itamar at ispbrasil.com.br  Mon Oct  5 08:48:23 2009
From: itamar at ispbrasil.com.br (Itamar Reis Peixoto)
Date: Mon, 5 Oct 2009 04:48:23 -0300
Subject: [Dnsmasq-discuss] sending patches do dnsmasq
Message-ID: <d3ceea380910050048q48d7b4a1r10d833a2812a4741@mail.gmail.com>

there are some patches in fedora, Can I send to this list ?

also can I send a patch to port dnsmasq to use gnu-autotools ?

gnu auto-tools is for

./configure
make
make install


-- 
------------

Itamar Reis Peixoto

e-mail/msn: itamar at ispbrasil.com.br
sip: itamar at ispbrasil.com.br
skype: itamarjp
icq: 81053601
+55 11 4063 5033
+55 34 3221 8599


From mark at jemms.net  Mon Oct  5 15:48:04 2009
From: mark at jemms.net (Mark Beierl)
Date: Mon, 05 Oct 2009 10:48:04 -0400
Subject: [Dnsmasq-discuss] dnsmasq as DHCP server on vmware client
	machine
In-Reply-To: <4AC810F5.4000102@tmtv.ne.jp>
References: <4AC810F5.4000102@tmtv.ne.jp>
Message-ID: <4ACA0724.9010605@jemms.net>

I use dnsmasq 2.50-1 with Vmware Workstation 6.5 and do not have any 
problems with vmnets.  The only difference I see is that I do not use 
bind-interfaces...

Regards,
Mark

Hisashi Hashimoto wrote:
> Hi,
>
>   I am using Dnsmasq 2.45, a little bit old on my Local distribution.
> I am trying to use dnsmasq as a DHCP server on vmware host machine
> instead of using vmware original DHCP server, vmnet-dhcpd.
> Why I prefer using dnsmasq is that vmnet-dhcpd is based on ISC DHCPD
> 2.0(old) and I prefer to use dnsmasq configuration.
>
>  I have one problem now.
> This proble appeared after I moved VMware workstation 6.5 from 5.0.
> Problem is:
> Dnsmasq does not look at the vmnet1, 4, 8(Virtual network on host
> machine created by VMware).
>
>   Why I think so, is :
> (1) When I use vmnet-dhcpd or ISC DHCPD 3.0, they answered the query
> from guest machines. but when I use dnsmasq, it did not answer.
>    So, network interfaces had been created and work without problems.
> (2) While I run ISC DHCPD, I started dnsmasq with DHCP-RANGE option,
>     Dnsmasq did not make any claim that port had been used by ISC DHCPD.
>
> Any suggestions are welcome.
> ---------------------------------------------------
> Here is my configuration related DHCP.
> interface=vmnet1
> interface=vmnet4
> interface=vmnet8
> interface=lo
> bind-interfaces
> dhcp-range=192.168.219.101,192.168.219.128,255.255.255.0,12h
> dhcp-range=192.168.140.101,192.168.140.254,255.255.255.0,12h
> dhcp-range=192.168.167.101,192.168.167.254,255.255.255.0,12h
>
> Below is output of ifconfig.
> vmnet1    Link encap:Ethernet  HWaddr 00:50:56:C0:00:01
>           inet addr:192.168.140.1  Bcast:192.168.140.255
>           Mask:255.255.255.0
>           inet6 addr: fe80::250:56ff:fec0:1/64 Scope:Link
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:1000
>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>
> vmnet4    Link encap:Ethernet  HWaddr 00:50:56:C0:00:04
>           inet addr:192.168.219.1  Bcast:192.168.219.255
>           Mask:255.255.255.0
>           inet6 addr: fe80::250:56ff:fec0:4/64 Scope:Link
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:25 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:1000
>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>
> vmnet8    Link encap:Ethernet  HWaddr 00:50:56:C0:00:08
>           inet addr:192.168.167.1  Bcast:192.168.167.255
>           Mask:255.255.255.0
>           inet6 addr: fe80::250:56ff:fec0:8/64 Scope:Link
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:0 txqueuelen:1000
>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>
> Hashimoto Hisashi
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>   


From simon at thekelleys.org.uk  Mon Oct  5 21:20:34 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 05 Oct 2009 21:20:34 +0100
Subject: [Dnsmasq-discuss] FAQ entry - Debian hostnames
In-Reply-To: <4AC5D04F.4040703@fundamentalsltd.co.uk>
References: <4AC5D04F.4040703@fundamentalsltd.co.uk>
Message-ID: <4ACA5512.90908@thekelleys.org.uk>

Roger Lynn wrote:
> Hi,
> 
> In the dnsmasq FAQ there's a question: "I'm running Debian, and my machines get
> an address fine with DHCP, but their names are not appearing in the DNS." The
> answer includes "Update: the lastest dhcpcd packages _do_ send the hostname by
> default." This isn't true for Debian, unless you are talking about the
> experimental packages of ISC DHCP 4.1.
> 
> A patch has apparently been applied in Ubuntu, but the Debian maintainers are
> waiting for upstream to fix it. See http://bugs.debian.org/151820 and
> http://bugs.debian.org/522889 for details.
> 
> Thanks for your work on dnsmasq,
> 

Unfortunate naming here. The FAQ entry refers to dhcpcd, which is a
completely different DHCP client than ISC offering, which is called
dhclient, I think. The statement in the FAQ is true for dhcpcd.


Cheers,

Simon.


From itamar at ispbrasil.com.br  Tue Oct  6 00:09:32 2009
From: itamar at ispbrasil.com.br (Itamar Reis Peixoto)
Date: Mon, 5 Oct 2009 20:09:32 -0300
Subject: [Dnsmasq-discuss] cosmetic patch use /var/lib/dnsmasq/ instead
	/var/lib/misc/
Message-ID: <d3ceea380910051609t6c80b9f7l9932a7a8f6313c8c@mail.gmail.com>

cosmetic patch use /var/lib/dnsmasq/ instead /var/lib/misc/

-#dhcp-leasefile=/var/lib/misc/dnsmasq.leases
+#dhcp-leasefile=/var/lib/dnsmasq/dnsmasq.leases

this patch is in fedora and I like to upstream it.


patch attached



-- 
------------

Itamar Reis Peixoto

e-mail/msn: itamar at ispbrasil.com.br
sip: itamar at ispbrasil.com.br
skype: itamarjp
icq: 81053601
+55 11 4063 5033
+55 34 3221 8599
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-configuration-cosmetic.patch
Type: application/octet-stream
Size: 1934 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091005/fcf745b9/attachment.obj 

From itamar at ispbrasil.com.br  Tue Oct  6 00:10:10 2009
From: itamar at ispbrasil.com.br (Itamar Reis Peixoto)
Date: Mon, 5 Oct 2009 20:10:10 -0300
Subject: [Dnsmasq-discuss] dnsmasq initscript for fedora/centos and other
	rpm distro.s
Message-ID: <d3ceea380910051610j234838dco6f6d31e327e2328f@mail.gmail.com>

patch attached


-- 
------------

Itamar Reis Peixoto

e-mail/msn: itamar at ispbrasil.com.br
sip: itamar at ispbrasil.com.br
skype: itamarjp
icq: 81053601
+55 11 4063 5033
+55 34 3221 8599
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-2.5.0-initscript.patch
Type: application/octet-stream
Size: 2251 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091005/53c495ef/attachment.obj 

From grexk at yahoo.com  Wed Oct  7 02:36:54 2009
From: grexk at yahoo.com (Greek Ordono)
Date: Tue, 6 Oct 2009 18:36:54 -0700 (PDT)
Subject: [Dnsmasq-discuss] dnsmasq dhcp-option 120
Message-ID: <867824.39501.qm@web53903.mail.re2.yahoo.com>

Hi

I getting segfaults when I try to set "dhcp-option=120,servername:port/cmcprov/login". I'm currently using dnsmasq-2.45-12.16.1 from openSUSE. TIA

--

Greek Ordono 

myppa: launchpad.net/~grexk/+archive/ppa


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091006/676cc840/attachment.htm 

From simon at thekelleys.org.uk  Wed Oct  7 10:11:53 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 07 Oct 2009 10:11:53 +0100
Subject: [Dnsmasq-discuss] dnsmasq dhcp-option 120
In-Reply-To: <867824.39501.qm@web53903.mail.re2.yahoo.com>
References: <867824.39501.qm@web53903.mail.re2.yahoo.com>
Message-ID: <4ACC5B59.3000708@thekelleys.org.uk>

Greek Ordono wrote:
> Hi
> 
> I getting segfaults when I try to set 
> "dhcp-option=120,servername:port/cmcprov/login". I'm currently using 
> dnsmasq-2.45-12.16.1 from openSUSE. TIA
> 
> --
> Greek Ordono
> myppa: launchpad.net/~grexk/+archive/ppa
> 
> 

Thanks for the report: this is fixed in later versions.


Cheers,

Simon.


From hashimoh at tmtv.ne.jp  Thu Oct  8 14:32:34 2009
From: hashimoh at tmtv.ne.jp (Hisashi Hashimoto)
Date: Thu, 08 Oct 2009 22:32:34 +0900
Subject: [Dnsmasq-discuss] dnsmasq as DHCP server on vmware
	client	machine
In-Reply-To: <4ACA0724.9010605@jemms.net>
References: <4AC810F5.4000102@tmtv.ne.jp> <4ACA0724.9010605@jemms.net>
Message-ID: <4ACDE9F2.70907@tmtv.ne.jp>

Mark,

   Thank you for your reply.
I comment out bind-interface, but nothing has been changed.

Hisashi

Mark Beierl wrote
> I use dnsmasq 2.50-1 with Vmware Workstation 6.5 and do not have any 
> problems with vmnets.  The only difference I see is that I do not use 
> bind-interfaces...
> 
> Regards,
> Mark
> 
> Hisashi Hashimoto wrote:
>> Hi,
>>
>>   I am using Dnsmasq 2.45, a little bit old on my Local distribution.
>> I am trying to use dnsmasq as a DHCP server on vmware host machine
>> instead of using vmware original DHCP server, vmnet-dhcpd.
>> Why I prefer using dnsmasq is that vmnet-dhcpd is based on ISC DHCPD
>> 2.0(old) and I prefer to use dnsmasq configuration.
>>
>>  I have one problem now.
>> This proble appeared after I moved VMware workstation 6.5 from 5.0.
>> Problem is:
>> Dnsmasq does not look at the vmnet1, 4, 8(Virtual network on host
>> machine created by VMware).
>>
>>   Why I think so, is :
>> (1) When I use vmnet-dhcpd or ISC DHCPD 3.0, they answered the query
>> from guest machines. but when I use dnsmasq, it did not answer.
>>    So, network interfaces had been created and work without problems.
>> (2) While I run ISC DHCPD, I started dnsmasq with DHCP-RANGE option,
>>     Dnsmasq did not make any claim that port had been used by ISC DHCPD.
>>
>> Any suggestions are welcome.
>> ---------------------------------------------------
>> Here is my configuration related DHCP.
>> interface=vmnet1
>> interface=vmnet4
>> interface=vmnet8
>> interface=lo
>> bind-interfaces
>> dhcp-range=192.168.219.101,192.168.219.128,255.255.255.0,12h
>> dhcp-range=192.168.140.101,192.168.140.254,255.255.255.0,12h
>> dhcp-range=192.168.167.101,192.168.167.254,255.255.255.0,12h
>>
>> Below is output of ifconfig.
>> vmnet1    Link encap:Ethernet  HWaddr 00:50:56:C0:00:01
>>           inet addr:192.168.140.1  Bcast:192.168.140.255
>>           Mask:255.255.255.0
>>           inet6 addr: fe80::250:56ff:fec0:1/64 Scope:Link
>>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>>           TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
>>           collisions:0 txqueuelen:1000
>>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>>
>> vmnet4    Link encap:Ethernet  HWaddr 00:50:56:C0:00:04
>>           inet addr:192.168.219.1  Bcast:192.168.219.255
>>           Mask:255.255.255.0
>>           inet6 addr: fe80::250:56ff:fec0:4/64 Scope:Link
>>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>>           TX packets:25 errors:0 dropped:0 overruns:0 carrier:0
>>           collisions:0 txqueuelen:1000
>>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>>
>> vmnet8    Link encap:Ethernet  HWaddr 00:50:56:C0:00:08
>>           inet addr:192.168.167.1  Bcast:192.168.167.255
>>           Mask:255.255.255.0
>>           inet6 addr: fe80::250:56ff:fec0:8/64 Scope:Link
>>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>>           RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>>           TX packets:24 errors:0 dropped:0 overruns:0 carrier:0
>>           collisions:0 txqueuelen:1000
>>           RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
>>
>> Hashimoto Hisashi
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>   
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 
> 


From bcook at poughkeepsieschools.org  Thu Oct  8 17:49:35 2009
From: bcook at poughkeepsieschools.org (B. Cook)
Date: Thu, 08 Oct 2009 12:49:35 -0400
Subject: [Dnsmasq-discuss] reload and restart; difference between?
Message-ID: <4ACE181F.3080606@poughkeepsieschools.org>

This might be a FreeBSD -only- question but..

Dnsmasq version 2.50  Copyright (C) 2000-2009 Simon Kelley
Compile time options no-IPv6 GNU-getopt no-DBus no-I18N DHCP TFTP


So I have a seemingly complex setup.. but suffice to say it all works as 
  designed..

I only hand out dhcp to known addresses; we recently had a deployment of 
300 new computers, which did not get entered into the "known hosts" file.

I did add the "known hostname" which allowed the pc to get an IP while 
its machine name matched the "known".  I made a little script that would 
grep the logs for the MAC addresses of the "known" hostname and add it 
to the "known hosts" file; and then restart the dnsmasq process (which 
was also doing dns).

This usually takes dnsmasq about 3-5 seconds to restart, which then 
causes people to be w/o dns (on a network of almost 3k machines) and 
makes the phone ring from time to time.. breaks cups in weird ways.. 
i.e. creates problems..

So I found that I could -reload- dnsmasq instead of -restart- it, but it 
seems that -reload- does not -reread- everything that has changed..

Is there someway to make this happen? or someway to get a -reload- to 
look at all the .conf and/or include files again?

Or is there something easier that I could do to accomplish what I am 
looking for?




From ourspamtrap at yahoo.com  Thu Oct  8 18:00:09 2009
From: ourspamtrap at yahoo.com (nitmd)
Date: Thu, 8 Oct 2009 10:00:09 -0700 (PDT)
Subject: [Dnsmasq-discuss] Local addresses don't come back as local
Message-ID: <150154.10237.qm@web53902.mail.re2.yahoo.com>

I'm working on installing dnsmasq for my small network.? I have it up and running, but the addresses I'm getting don't make sense.

My network is 192.168.27.?? I have a machine named fred.? If I ping fred, I get a response

Reply from 216.146.8.16:? etc etc

I'm wondering what I missed in setup, as it should come back from 192.168.27.xx




      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091008/95da580d/attachment.htm 

From ourspamtrap at yahoo.com  Thu Oct  8 18:05:17 2009
From: ourspamtrap at yahoo.com (nitmd)
Date: Thu, 8 Oct 2009 10:05:17 -0700 (PDT)
Subject: [Dnsmasq-discuss] local dns not local [solved??]
Message-ID: <19741.13407.qm@web53912.mail.re2.yahoo.com>

I found part of the problem, at least; my windows machine reverted to an older dns server setting for some reason and was ignoring the local machine.? I'll see if this happens again and deal with it then.



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091008/082a996e/attachment.htm 

From ourspamtrap at yahoo.com  Thu Oct  8 21:11:23 2009
From: ourspamtrap at yahoo.com (nitmd)
Date: Thu, 8 Oct 2009 13:11:23 -0700 (PDT)
Subject: [Dnsmasq-discuss] DNS server changing -solved
Message-ID: <46756.46425.qm@web53911.mail.re2.yahoo.com>

I was having trouble with my dnsmasq server being ignored, discovered it was because my windows machine was changing my dns server addresses without my knowledge.? I found out this is because I am using the dyndns updater, and had the "internet guide" enabled.? This apparently causes dyndns to change your server settings to use their server, eliminating the settings you have put in.? If you use that updater, watch for the option to enable/disable this service when you install.



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091008/21adc163/attachment.htm 

From david at howes.net.nz  Thu Oct  8 23:29:11 2009
From: david at howes.net.nz (David Howes)
Date: Fri, 9 Oct 2009 11:29:11 +1300
Subject: [Dnsmasq-discuss] Setting to always serve from /etc/hosts,
	even when upstream servers 	are down.
Message-ID: <b2705cf90910081529j3b2ca6dcy1393ec974c7c8bfa@mail.gmail.com>

Hi!

I have DNSMasq in a local network, we have one issue.

When the upstream servers are down DNSMasq won't even serve entries from the
/etc/hosts file.
I need DNSMasq always to serve from the /etc/hosts file if an entry exists,
even if the upstream servers are down.

Is this possible with DNSMasq currently, if not, can it be added?

Thanks in advance,
David Howes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091009/580cc6f8/attachment.htm 

From richardvoigt at gmail.com  Fri Oct  9 01:54:44 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Thu, 8 Oct 2009 19:54:44 -0500
Subject: [Dnsmasq-discuss] reload and restart; difference between?
In-Reply-To: <4ACE181F.3080606@poughkeepsieschools.org>
References: <4ACE181F.3080606@poughkeepsieschools.org>
Message-ID: <2e59e6970910081754t378517esbe8594d956929b79@mail.gmail.com>

Use a separate hostsfile, which is read during reload.

The configuration file cannot be reprocessed without a complete
restart because dnsmasq drops root early on and no longer has the
permission needed to listen on privileged ports..

On Thu, Oct 8, 2009 at 11:49 AM, B. Cook <bcook at poughkeepsieschools.org> wrote:
> This might be a FreeBSD -only- question but..
>
> Dnsmasq version 2.50 ?Copyright (C) 2000-2009 Simon Kelley
> Compile time options no-IPv6 GNU-getopt no-DBus no-I18N DHCP TFTP
>
>
> So I have a seemingly complex setup.. but suffice to say it all works as
> ?designed..
>
> I only hand out dhcp to known addresses; we recently had a deployment of
> 300 new computers, which did not get entered into the "known hosts" file.
>
> I did add the "known hostname" which allowed the pc to get an IP while
> its machine name matched the "known". ?I made a little script that would
> grep the logs for the MAC addresses of the "known" hostname and add it
> to the "known hosts" file; and then restart the dnsmasq process (which
> was also doing dns).
>
> This usually takes dnsmasq about 3-5 seconds to restart, which then
> causes people to be w/o dns (on a network of almost 3k machines) and
> makes the phone ring from time to time.. breaks cups in weird ways..
> i.e. creates problems..
>
> So I found that I could -reload- dnsmasq instead of -restart- it, but it
> seems that -reload- does not -reread- everything that has changed..
>
> Is there someway to make this happen? or someway to get a -reload- to
> look at all the .conf and/or include files again?
>
> Or is there something easier that I could do to accomplish what I am
> looking for?
>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eldonz at atlanticdb.com  Sat Oct 10 17:57:46 2009
From: eldonz at atlanticdb.com (Eldon Ziegler)
Date: Sat, 10 Oct 2009 12:57:46 -0400
Subject: [Dnsmasq-discuss] Use Fedora 11 kickstart with dnsmasq?
Message-ID: <1255193866.2926.423.camel@carmen.atlantic>

Simon,

Is there a way to have dnsmasq give a Fedora 11 install the location of
a kickstart file?

From
http://docs.fedoraproject.org/install-guide/f11/en-US/html/s2-kickstart2-networkbased.html

Here is an example of a line from the dhcpd.conf file for the DHCP server: 

filename "/usr/new-machine/kickstart/"; next-server blarg.redhat.com;

Note that you should replace the value after filename with the name of
the kickstart file (or the directory in which the kickstart file
resides) and the value after next-server with the NFS server name. 

Thanks,
Eldon




From richardvoigt at gmail.com  Sat Oct 10 19:17:20 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sat, 10 Oct 2009 13:17:20 -0500
Subject: [Dnsmasq-discuss] Use Fedora 11 kickstart with dnsmasq?
In-Reply-To: <1255193866.2926.423.camel@carmen.atlantic>
References: <1255193866.2926.423.camel@carmen.atlantic>
Message-ID: <2e59e6970910101117v785e0de5j163b6a9e856f17ff@mail.gmail.com>

dnsmasq can give out NFS file information, it's needed for BOOTP.

You'll want to make sure to mark the configuration so the kickstart
information only goes out to fedora-class clients.  Having a BOOTP
BIOS try to load your kickstart as a kernel is probably not going to
have the results you want.

On Sat, Oct 10, 2009 at 11:57 AM, Eldon Ziegler <eldonz at atlanticdb.com> wrote:
> Simon,
>
> Is there a way to have dnsmasq give a Fedora 11 install the location of
> a kickstart file?
>
> From
> http://docs.fedoraproject.org/install-guide/f11/en-US/html/s2-kickstart2-networkbased.html
>
> Here is an example of a line from the dhcpd.conf file for the DHCP server:
>
> filename "/usr/new-machine/kickstart/"; next-server blarg.redhat.com;
>
> Note that you should replace the value after filename with the name of
> the kickstart file (or the directory in which the kickstart file
> resides) and the value after next-server with the NFS server name.
>
> Thanks,
> Eldon
>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eldonz at atlanticdb.com  Sun Oct 11 12:42:38 2009
From: eldonz at atlanticdb.com (Eldon Ziegler)
Date: Sun, 11 Oct 2009 07:42:38 -0400
Subject: [Dnsmasq-discuss] Use Fedora 11 kickstart with dnsmasq?
In-Reply-To: <2e59e6970910101117v785e0de5j163b6a9e856f17ff@mail.gmail.com>
References: <1255193866.2926.423.camel@carmen.atlantic>
	<2e59e6970910101117v785e0de5j163b6a9e856f17ff@mail.gmail.com>
Message-ID: <1255261358.2926.429.camel@carmen.atlantic>

Thanks for the information. What should be added to the dnsmasq.conf
file to make this work?


On Sat, 2009-10-10 at 13:17 -0500, richardvoigt at gmail.com wrote:
> dnsmasq can give out NFS file information, it's needed for BOOTP.
> 
> You'll want to make sure to mark the configuration so the kickstart
> information only goes out to fedora-class clients.  Having a BOOTP
> BIOS try to load your kickstart as a kernel is probably not going to
> have the results you want.
> 
> On Sat, Oct 10, 2009 at 11:57 AM, Eldon Ziegler <eldonz at atlanticdb.com> wrote:
> > Simon,
> >
> > Is there a way to have dnsmasq give a Fedora 11 install the location of
> > a kickstart file?
> >
> > From
> > http://docs.fedoraproject.org/install-guide/f11/en-US/html/s2-kickstart2-networkbased.html
> >
> > Here is an example of a line from the dhcpd.conf file for the DHCP server:
> >
> > filename "/usr/new-machine/kickstart/"; next-server blarg.redhat.com;
> >
> > Note that you should replace the value after filename with the name of
> > the kickstart file (or the directory in which the kickstart file
> > resides) and the value after next-server with the NFS server name.
> >
> > Thanks,
> > Eldon
> >
> >
> >
> > _______________________________________________
> > Dnsmasq-discuss mailing list
> > Dnsmasq-discuss at lists.thekelleys.org.uk
> > http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> >



From clintonlee.taylor at gmail.com  Sun Oct 11 14:26:34 2009
From: clintonlee.taylor at gmail.com (Clinton Lee Taylor)
Date: Sun, 11 Oct 2009 15:26:34 +0200
Subject: [Dnsmasq-discuss] Multiple resolv.conf files ...
Message-ID: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>

Greetings ...

 Got to thank Simon for some really great work ... DnsMasq has made my
life so much easier for so long, thanks does not seem enough ...

 Been looking into creating a Linux Router that can be used for
internet connection fall over, when the primary internet connection
fails, I would like to have the router switch to the backup.  I know
that DnsMasq can't really help with this, but, I'm looking at making a
second connection, using 3G ...

 So, my problem is that without making very complicated scripts, could
we not make DnsMasq be able to read multiple resolv.conf type files
...

 The reason I would like this, is that each internet facing interface,
using something like DHCP, would have their own DNS server list that I
could get DnsMasq to read/use, so that DNS queries would be sent down
each active internet interface ...

Thanks
Mailed
LeeT


From richardvoigt at gmail.com  Sun Oct 11 20:04:38 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 11 Oct 2009 14:04:38 -0500
Subject: [Dnsmasq-discuss] Use Fedora 11 kickstart with dnsmasq?
In-Reply-To: <1255261358.2926.429.camel@carmen.atlantic>
References: <1255193866.2926.423.camel@carmen.atlantic>
	<2e59e6970910101117v785e0de5j163b6a9e856f17ff@mail.gmail.com>
	<1255261358.2926.429.camel@carmen.atlantic>
Message-ID: <2e59e6970910111204l7e4f8193x7f2c06fc940099c9@mail.gmail.com>

Sorry for the misinformation.  BOOTP support in dnsmasq provides a
TFTP server.  Are you sure the reference to NFS in the kickstart docs
shouldn't really be TFTP?  In any case, the BOOTP RFC seems to suggest
that alternate protocols are possible.  Try using the dhcp-boot
directive in your dnsmasq configuration.  Also try reading the
excellent man page provided by Simon (dnsmasq's author).

On Sun, Oct 11, 2009 at 6:42 AM, Eldon Ziegler <eldonz at atlanticdb.com> wrote:
> Thanks for the information. What should be added to the dnsmasq.conf
> file to make this work?
>
>
> On Sat, 2009-10-10 at 13:17 -0500, richardvoigt at gmail.com wrote:
>> dnsmasq can give out NFS file information, it's needed for BOOTP.
>>
>> You'll want to make sure to mark the configuration so the kickstart
>> information only goes out to fedora-class clients. ?Having a BOOTP
>> BIOS try to load your kickstart as a kernel is probably not going to
>> have the results you want.
>>
>> On Sat, Oct 10, 2009 at 11:57 AM, Eldon Ziegler <eldonz at atlanticdb.com> wrote:
>> > Simon,
>> >
>> > Is there a way to have dnsmasq give a Fedora 11 install the location of
>> > a kickstart file?
>> >
>> > From
>> > http://docs.fedoraproject.org/install-guide/f11/en-US/html/s2-kickstart2-networkbased.html
>> >
>> > Here is an example of a line from the dhcpd.conf file for the DHCP server:
>> >
>> > filename "/usr/new-machine/kickstart/"; next-server blarg.redhat.com;
>> >
>> > Note that you should replace the value after filename with the name of
>> > the kickstart file (or the directory in which the kickstart file
>> > resides) and the value after next-server with the NFS server name.
>> >
>> > Thanks,
>> > Eldon
>> >
>> >
>> >
>> > _______________________________________________
>> > Dnsmasq-discuss mailing list
>> > Dnsmasq-discuss at lists.thekelleys.org.uk
>> > http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>> >
>
>


From richardvoigt at gmail.com  Sun Oct 11 20:07:38 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 11 Oct 2009 14:07:38 -0500
Subject: [Dnsmasq-discuss] Multiple resolv.conf files ...
In-Reply-To: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>
References: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>
Message-ID: <2e59e6970910111207w7672041bt9c7dd932f9adb467@mail.gmail.com>

On Sun, Oct 11, 2009 at 8:26 AM, Clinton Lee Taylor
<clintonlee.taylor at gmail.com> wrote:
> Greetings ...
>
> ?Got to thank Simon for some really great work ... DnsMasq has made my
> life so much easier for so long, thanks does not seem enough ...
>
> ?Been looking into creating a Linux Router that can be used for
> internet connection fall over, when the primary internet connection
> fails, I would like to have the router switch to the backup. ?I know
> that DnsMasq can't really help with this, but, I'm looking at making a
> second connection, using 3G ...
>
> ?So, my problem is that without making very complicated scripts, could
> we not make DnsMasq be able to read multiple resolv.conf type files

Simon already has this covered, read your man page, which says
"Dnsmasq can be told to poll more than one resolv.conf file".

http://www.thekelleys.org.uk/dnsmasq/docs/dnsmasq-man.html

You'll solve your problem a lot faster than waiting for an answer from the list.

> ...
>
> ?The reason I would like this, is that each internet facing interface,
> using something like DHCP, would have their own DNS server list that I
> could get DnsMasq to read/use, so that DNS queries would be sent down
> each active internet interface ...
>
> Thanks
> Mailed
> LeeT
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From simon at thekelleys.org.uk  Tue Oct 13 18:04:56 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Tue, 13 Oct 2009 18:04:56 +0100
Subject: [Dnsmasq-discuss] Announce: dnsmasq-2.51
Message-ID: <4AD4B338.7080206@thekelleys.org.uk>

Dnsmasq 2.51 is now available from

http://www.thekelleys.org.uk/dnsmasq/dnsmasq-2.51.tar.gz

The Changelog looks like this:

            Add support for internationalised DNS. Non-ASCII characters
            in domain names found in /etc/hosts, /etc/ethers and
	    /etc/dnsmasq.conf will be correctly handled by translation
            to punycode, as specified in RFC3490. This function is only
            available if dnsmasq is compiled with internationalisation
            support, and adds a dependency on GNU libidn. Without i18n
            support, dnsmasq continues to be compilable with just
            standard tools. Thanks to Yves Dorfsman for the
            suggestion.

            Add two more environment variables for lease-change scripts:
	    First, DNSMASQ_SUPPLIED_HOSTNAME; this is set to the
            hostname supplied by a client, even if the actual hostname
            used is over-ridden by dhcp-host or dhcp-ignore-names
            directives. Also DNSMASQ_RELAY_ADDRESS which gives the
            address of a DHCP relay, if used.
	    Suggestions from Michael Rack.

	    Fix regression which broke echo of relay-agent
	    options. Thanks to Michael Rack for spotting this.

            Don't treat option 67 as being interchangeable with
            dhcp-boot parameters if it's specified as
            dhcp-option-force.

	    Make the code to call scripts on lease-change compile-time
	    optional. It can be switched off by editing src/config.h
	    or building with "make COPTS=-DNO_SCRIPT".

	    Make the TFTP server cope with filenames from Windows/DOS
	    which use '\' as pathname separator. Thanks to Ralf for
	    the patch.

	    Updated Polish translation. Thanks to Jan Psota.

	    Warn if an IP address is duplicated in /etc/ethers. Thanks
	    to Felix Schwarz for pointing this out.

	    Teach --conf-dir to take an option list of file suffices
	    which will be ignored when scanning the directory. Useful
	    for backup files etc. Thanks to Helmut Hullen for the
	    suggestion.

	    Add new DHCP option named tftpserver-address, which
	    corresponds to the third argument of dhcp-boot. This
	    allows the complete functionality of dhcp-boot to be
	    replicated with dhcp-option. Useful when using
	    dhcp-optsfile.

	    Test which upstream nameserver to use every 10 seconds
            or 50 queries and not just when a query times out and
            is retried. This should improve performance when there
            is a slow nameserver in the list. Thanks to Joe for the
            suggestion.

	    Don't do any PXE processing, even for clients with the
	    correct vendorclass, unless at least one pxe-prompt or
            pxe-service option is given. This stops dnsmasq
            interfering with proxy PXE subsystems when it is just
            the DHCP server. Thanks to Spencer Clark for spotting this.

	    Limit the blocksize used for TFTP transfers to a value
	    which avoids packet fragmentation, based on the MTU of the
	    local interface. Many netboot ROMs can't cope with
	    fragmented packets.

	    Honour dhcp-ignore configuration for PXE and proxy-PXE
	    requests. Thanks to Niels Basjes for the bug report.

            Updated French translation. Thanks to Gildas Le Nadan.


Cheers,

Simon.


From tommy at brander.nu  Wed Oct 14 08:29:07 2009
From: tommy at brander.nu (Tommy Brander)
Date: Wed, 14 Oct 2009 09:29:07 +0200
Subject: [Dnsmasq-discuss] Problems with custom dhcp-script and leasefile-ro
Message-ID: <1255505347.5144.0.camel@pisces>

Hi All

I'm using dnsmasq version 2.47 on Ubuntu Server Jaunty with a custom
dhcp-script and the option leasefile-ro. The script is a PHP file which
stores the lease database together with available DHCP options (vendorid
and so on) in a PostgreSQL database.
After a restart of dnsmasq, when the script is called with init, i read
the active leases and print them with the following statement:

fputs(STDOUT, "$lease");

where $lease = '1255470657 00:1f:29:b4:84:82 192.168.0.194 pisces *'.
Observe that this is without any added \n.

I also log this to a log file, that after init contains the following:

2009-10-13 23:23:18 - init: 1255470657 00:1f:29:b4:84:82 192.168.0.194
pisces *
2009-10-13 23:23:18 - init: 1255470430 00:01:38:99:52:80 192.168.0.182
betelgeuse 01:00:01:38:99:52:80
2009-10-13 23:23:18 - init: 1255470317 00:21:00:d2:25:4e 192.168.0.113
Delphinus 01:00:21:00:d2:25:4e
2009-10-13 23:23:18 - init: 1255470271 08:00:27:23:dd:64 192.168.0.104
caelum 01:08:00:27:23:dd:64
2009-10-13 23:23:18 - init: 1255470066 00:02:61:0f:c2:ec 192.168.0.123 *
01:00:02:61:0f:c2:ec

In the syslog, the following rows has appeared:

Oct 13 23:23:17 aurora dnsmasq[5420]: exiting on receipt of SIGTERM
Oct 13 23:23:18 aurora dnsmasq[5611]: started, version 2.47 cachesize
150
Oct 13 23:23:18 aurora dnsmasq[5611]: compile time options: IPv6
GNU-getopt DBus I18N TFTP
Oct 13 23:23:18 aurora dnsmasq[5611]: DHCP, IP range 192.168.0.101 --
192.168.0.199, lease time 30m
Oct 13 23:23:18 aurora dnsmasq[5611]: reading /etc/resolv.conf
Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.200#53
Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.198#53
Oct 13 23:23:18 aurora dnsmasq[5611]: read /etc/hosts - 21 addresses

The problem I'm having with this setup is that when the host pisces,
which is a Ubuntu Karmic host, request it's next IP, the event generated
is not an 'Old' but an 'Add'. This is the row that appears in my dhcp
log file:

2009-10-13 23:23:45 - add: 00:1f:29:b4:84:82 192.168.0.194 pisces  eth1
1800  brander.local

Is that correct? The client first does two DHCPREQUEST, which dnsmasq
host responds with DHCPNAK. Then the client does a DHCPDISCOVER and gets
an IP - some times the same IP, some times a different IP. Is this a
feature or a bug?

Since dnsmasq read the lease file during init, it should give out the
same address and generate an 'Old' event. Or am I just stupid?

Can anyone help me?


Best Regards, Tommy Brander
tommy at brander.nu  ? www.brander.nu

Viam inveniam aut faciam - I shall either find a way or make one 

Best Regards, Tommy Brander
tommy at brander.nu  ? www.brander.nu

Viam inveniam aut faciam - I shall either find a way or make one 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091014/5a18ed9f/attachment.htm 

From simon at thekelleys.org.uk  Wed Oct 14 09:07:34 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 14 Oct 2009 09:07:34 +0100
Subject: [Dnsmasq-discuss] Problems with custom dhcp-script and
	leasefile-ro
In-Reply-To: <1255505347.5144.0.camel@pisces>
References: <1255505347.5144.0.camel@pisces>
Message-ID: <4AD586C6.2050709@thekelleys.org.uk>

Tommy Brander wrote:
> Hi All
> 
> I'm using dnsmasq version 2.47 on Ubuntu Server Jaunty with a custom 
> dhcp-script and the option leasefile-ro. The script is a PHP file which 
> stores the lease database together with available DHCP options (vendorid 
> and so on) in a PostgreSQL database.
> After a restart of dnsmasq, when the script is called with init, i read 
> the active leases and print them with the following statement:
> 
> fputs(STDOUT, "$lease");
> 
> where $lease = '1255470657 00:1f:29:b4:84:82 192.168.0.194 pisces *'. 
> Observe that this is without any added \n.
> 
> I also log this to a log file, that after init contains the following:
> 
> 2009-10-13 23:23:18 - init: 1255470657 00:1f:29:b4:84:82 192.168.0.194 
> pisces *
> 2009-10-13 23:23:18 - init: 1255470430 00:01:38:99:52:80 192.168.0.182 
> betelgeuse 01:00:01:38:99:52:80
> 2009-10-13 23:23:18 - init: 1255470317 00:21:00:d2:25:4e 192.168.0.113 
> Delphinus 01:00:21:00:d2:25:4e
> 2009-10-13 23:23:18 - init: 1255470271 08:00:27:23:dd:64 192.168.0.104 
> caelum 01:08:00:27:23:dd:64
> 2009-10-13 23:23:18 - init: 1255470066 00:02:61:0f:c2:ec 192.168.0.123 * 
> 01:00:02:61:0f:c2:ec
> 
> In the syslog, the following rows has appeared:
> 
> Oct 13 23:23:17 aurora dnsmasq[5420]: exiting on receipt of SIGTERM
> Oct 13 23:23:18 aurora dnsmasq[5611]: started, version 2.47 cachesize 150
> Oct 13 23:23:18 aurora dnsmasq[5611]: compile time options: IPv6 
> GNU-getopt DBus I18N TFTP
> Oct 13 23:23:18 aurora dnsmasq[5611]: DHCP, IP range 192.168.0.101 -- 
> 192.168.0.199, lease time 30m
> Oct 13 23:23:18 aurora dnsmasq[5611]: reading /etc/resolv.conf
> Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.200#53
> Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.198#53
> Oct 13 23:23:18 aurora dnsmasq[5611]: read /etc/hosts - 21 addresses
> 
> The problem I'm having with this setup is that when the host pisces, 
> which is a Ubuntu Karmic host, request it's next IP, the event generated 
> is not an 'Old' but an 'Add'. This is the row that appears in my dhcp 
> log file:
> 
> 2009-10-13 23:23:45 - add: 00:1f:29:b4:84:82 192.168.0.194 pisces  eth1 
> 1800  brander.local
> 
> Is that correct? The client first does two DHCPREQUEST, which dnsmasq 
> host responds with DHCPNAK. Then the client does a DHCPDISCOVER and gets 
> an IP - some times the same IP, some times a different IP. Is this a 
> feature or a bug?
> 
> Since dnsmasq read the lease file during init, it should give out the 
> same address and generate an 'Old' event. Or am I just stupid?
> 
> Can anyone help me?
> 

I guess the first step is to find out exactly what leases dnsmasq has in 
its database after the "init" event. The easiest way to do this is to 
run dnsmasq with the --no-daemon --log-queries flags, and then send it a 
USR1 signal. That will dump the DNS cache to the stdout. DNS entries 
which arise from DHCP leases will have a "D" in the flags field. You 
should see one for each of your DHCP leases which have a name.

My guess, without looking in detail, is that you have already found the 
problem - the lack of newline characters.

HTH

Simon.


From dnsmasq.list at michaelrack.de  Wed Oct 14 09:47:20 2009
From: dnsmasq.list at michaelrack.de (Michael Rack)
Date: Wed, 14 Oct 2009 10:47:20 +0200
Subject: [Dnsmasq-discuss] Problems with custom dhcp-script
	and	leasefile-ro
In-Reply-To: <4AD586C6.2050709@thekelleys.org.uk>
References: <1255505347.5144.0.camel@pisces>
	<4AD586C6.2050709@thekelleys.org.uk>
Message-ID: <4AD59018.5070309@michaelrack.de>

Of course, the newline character ist the problem :-D

fwrite is a binary-safe file write function. You have to replace your 
code-line with:

     fputs(STDOUT, $lease . chr(10));

And please .. PLEASE .. remove the double-quotes! What to the hell 
should the double-quotes make sense? Double-Quotes in PHP make the 
script execution slower, because PHP have to handle 
inline-defined-variables.

Cheers,
Michael.

Am 14.10.2009 10:07, schrieb Simon Kelley:
> Tommy Brander wrote:
>    
>> Hi All
>>
>> I'm using dnsmasq version 2.47 on Ubuntu Server Jaunty with a custom
>> dhcp-script and the option leasefile-ro. The script is a PHP file which
>> stores the lease database together with available DHCP options (vendorid
>> and so on) in a PostgreSQL database.
>> After a restart of dnsmasq, when the script is called with init, i read
>> the active leases and print them with the following statement:
>>
>> fputs(STDOUT, "$lease");
>>
>> where $lease = '1255470657 00:1f:29:b4:84:82 192.168.0.194 pisces *'.
>> Observe that this is without any added \n.
>>
>> I also log this to a log file, that after init contains the following:
>>
>> 2009-10-13 23:23:18 - init: 1255470657 00:1f:29:b4:84:82 192.168.0.194
>> pisces *
>> 2009-10-13 23:23:18 - init: 1255470430 00:01:38:99:52:80 192.168.0.182
>> betelgeuse 01:00:01:38:99:52:80
>> 2009-10-13 23:23:18 - init: 1255470317 00:21:00:d2:25:4e 192.168.0.113
>> Delphinus 01:00:21:00:d2:25:4e
>> 2009-10-13 23:23:18 - init: 1255470271 08:00:27:23:dd:64 192.168.0.104
>> caelum 01:08:00:27:23:dd:64
>> 2009-10-13 23:23:18 - init: 1255470066 00:02:61:0f:c2:ec 192.168.0.123 *
>> 01:00:02:61:0f:c2:ec
>>
>> In the syslog, the following rows has appeared:
>>
>> Oct 13 23:23:17 aurora dnsmasq[5420]: exiting on receipt of SIGTERM
>> Oct 13 23:23:18 aurora dnsmasq[5611]: started, version 2.47 cachesize 150
>> Oct 13 23:23:18 aurora dnsmasq[5611]: compile time options: IPv6
>> GNU-getopt DBus I18N TFTP
>> Oct 13 23:23:18 aurora dnsmasq[5611]: DHCP, IP range 192.168.0.101 --
>> 192.168.0.199, lease time 30m
>> Oct 13 23:23:18 aurora dnsmasq[5611]: reading /etc/resolv.conf
>> Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.200#53
>> Oct 13 23:23:18 aurora dnsmasq[5611]: using nameserver 195.54.122.198#53
>> Oct 13 23:23:18 aurora dnsmasq[5611]: read /etc/hosts - 21 addresses
>>
>> The problem I'm having with this setup is that when the host pisces,
>> which is a Ubuntu Karmic host, request it's next IP, the event generated
>> is not an 'Old' but an 'Add'. This is the row that appears in my dhcp
>> log file:
>>
>> 2009-10-13 23:23:45 - add: 00:1f:29:b4:84:82 192.168.0.194 pisces  eth1
>> 1800  brander.local
>>
>> Is that correct? The client first does two DHCPREQUEST, which dnsmasq
>> host responds with DHCPNAK. Then the client does a DHCPDISCOVER and gets
>> an IP - some times the same IP, some times a different IP. Is this a
>> feature or a bug?
>>
>> Since dnsmasq read the lease file during init, it should give out the
>> same address and generate an 'Old' event. Or am I just stupid?
>>
>> Can anyone help me?
>>
>>      
> I guess the first step is to find out exactly what leases dnsmasq has in
> its database after the "init" event. The easiest way to do this is to
> run dnsmasq with the --no-daemon --log-queries flags, and then send it a
> USR1 signal. That will dump the DNS cache to the stdout. DNS entries
> which arise from DHCP leases will have a "D" in the flags field. You
> should see one for each of your DHCP leases which have a name.
>
> My guess, without looking in detail, is that you have already found the
> problem - the lack of newline characters.
>
> HTH
>
> Simon.
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>    



From clintonlee.taylor at gmail.com  Wed Oct 14 11:51:54 2009
From: clintonlee.taylor at gmail.com (Clinton Lee Taylor)
Date: Wed, 14 Oct 2009 12:51:54 +0200
Subject: [Dnsmasq-discuss] Multiple resolv.conf files ...
In-Reply-To: <2e59e6970910111207w7672041bt9c7dd932f9adb467@mail.gmail.com>
References: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>
	<2e59e6970910111207w7672041bt9c7dd932f9adb467@mail.gmail.com>
Message-ID: <b7b14cbb0910140351i22e0464i2882dda9323768f@mail.gmail.com>

Greetings ...

>> ?Been looking into creating a Linux Router that can be used for
>> internet connection fall over, when the primary internet connection
>> fails, I would like to have the router switch to the backup. ?I know
>> that DnsMasq can't really help with this, but, I'm looking at making a
>> second connection, using 3G ...
>>
>> ?So, my problem is that without making very complicated scripts, could
>> we not make DnsMasq be able to read multiple resolv.conf type files
>
> Simon already has this covered, read your man page, which says
> "Dnsmasq can be told to poll more than one resolv.conf file".
 Arh, yes ... I did see this and hoped that it would work similar to
other multiple options ...

> http://www.thekelleys.org.uk/dnsmasq/docs/dnsmasq-man.html
 Went through that, but I did not get it to work as I had hoped ...

> You'll solve your problem a lot faster than waiting for an answer from the list.
 Mmm, right and wrong ... If you read the little bit before, I was
hoping that I could read multiple resolv.conf type files, but testing
showed something that did not seem to be working as I had expected ...
Was hoping that either what I was epxecting, could be helped or maybe
another insight into another way to do a similar thing ...

>> ?The reason I would like this, is that each internet facing interface,
>> using something like DHCP, would have their own DNS server list that I
>> could get DnsMasq to read/use, so that DNS queries would be sent down
>> each active internet interface ...

 I added to my dnsmasq.conf file two lines as follows ...

resolv-file=/etc/resolv.conf.dnsmasq-eth0
resolv-file=/etc/resolv.conf.dnsmasq-3g

 But when I restart dnsmasq, it only list the last resolv.conf file
and the dns server there in ... Do I need to use the -r option from
the command line ( init script ) or can this option be used in the
dnsmasq.conf file?

Oct 14 12:49:30 hermes dnsmasq[8652]: exiting on receipt of SIGTERM
Oct 14 12:49:30 hermes dnsmasq[8700]: started, version 2.45 cachesize 150
Oct 14 12:49:30 hermes dnsmasq[8700]: compile time options: IPv6
GNU-getopt no-ISC-leasefile no-DBus no-I18N TFTP
Oct 14 12:49:30 hermes dnsmasq[8700]: DHCP, IP range 192.168.1.65 --
192.168.1.115, lease time 4h
Oct 14 12:49:30 hermes dnsmasq[8700]: reading /etc/resolv.conf.dnsmasq-3g
Oct 14 12:49:30 hermes dnsmasq[8700]: using nameserver 208.67.222.222#53
Oct 14 12:49:30 hermes dnsmasq[8700]: read /etc/hosts - 8 addresses

Thanks
Mailed
LeeT


From simon at thekelleys.org.uk  Wed Oct 14 12:03:28 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 14 Oct 2009 12:03:28 +0100
Subject: [Dnsmasq-discuss] Multiple resolv.conf files ...
In-Reply-To: <b7b14cbb0910140351i22e0464i2882dda9323768f@mail.gmail.com>
References: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>	<2e59e6970910111207w7672041bt9c7dd932f9adb467@mail.gmail.com>
	<b7b14cbb0910140351i22e0464i2882dda9323768f@mail.gmail.com>
Message-ID: <4AD5B000.1030708@thekelleys.org.uk>

Clinton Lee Taylor wrote:
> Greetings ...
> 
>>>  Been looking into creating a Linux Router that can be used for
>>> internet connection fall over, when the primary internet connection
>>> fails, I would like to have the router switch to the backup.  I know
>>> that DnsMasq can't really help with this, but, I'm looking at making a
>>> second connection, using 3G ...
>>>
>>>  So, my problem is that without making very complicated scripts, could
>>> we not make DnsMasq be able to read multiple resolv.conf type files
>> Simon already has this covered, read your man page, which says
>> "Dnsmasq can be told to poll more than one resolv.conf file".
>  Arh, yes ... I did see this and hoped that it would work similar to
> other multiple options ...
> 
>> http://www.thekelleys.org.uk/dnsmasq/docs/dnsmasq-man.html
>  Went through that, but I did not get it to work as I had hoped ...
> 
>> You'll solve your problem a lot faster than waiting for an answer from the list.
>  Mmm, right and wrong ... If you read the little bit before, I was
> hoping that I could read multiple resolv.conf type files, but testing
> showed something that did not seem to be working as I had expected ...
> Was hoping that either what I was epxecting, could be helped or maybe
> another insight into another way to do a similar thing ...
> 
>>>  The reason I would like this, is that each internet facing interface,
>>> using something like DHCP, would have their own DNS server list that I
>>> could get DnsMasq to read/use, so that DNS queries would be sent down
>>> each active internet interface ...
> 
>  I added to my dnsmasq.conf file two lines as follows ...
> 
> resolv-file=/etc/resolv.conf.dnsmasq-eth0
> resolv-file=/etc/resolv.conf.dnsmasq-3g
> 
>  But when I restart dnsmasq, it only list the last resolv.conf file
> and the dns server there in ... Do I need to use the -r option from
> the command line ( init script ) or can this option be used in the
> dnsmasq.conf file?
> 
> Oct 14 12:49:30 hermes dnsmasq[8652]: exiting on receipt of SIGTERM
> Oct 14 12:49:30 hermes dnsmasq[8700]: started, version 2.45 cachesize 150
> Oct 14 12:49:30 hermes dnsmasq[8700]: compile time options: IPv6
> GNU-getopt no-ISC-leasefile no-DBus no-I18N TFTP
> Oct 14 12:49:30 hermes dnsmasq[8700]: DHCP, IP range 192.168.1.65 --
> 192.168.1.115, lease time 4h
> Oct 14 12:49:30 hermes dnsmasq[8700]: reading /etc/resolv.conf.dnsmasq-3g
> Oct 14 12:49:30 hermes dnsmasq[8700]: using nameserver 208.67.222.222#53
> Oct 14 12:49:30 hermes dnsmasq[8700]: read /etc/hosts - 8 addresses
> 
> Thanks
> Mailed
> LeeT

Dnsmasq checks the modification times of both (all) the files, and uses 
the one that was changed last. Try this:

touch /etc/resolv.conf.dnsmasq-eth0
dig www.google.com

You should see dnsmasq read the eth0 file and start to use the 
nameservers in there.

You need to ensure that the relevant resolv.conf file is updated when a 
network link goes live and all should be well.

Cheers,

Simon.



From clintonlee.taylor at gmail.com  Wed Oct 14 13:27:09 2009
From: clintonlee.taylor at gmail.com (Clinton Lee Taylor)
Date: Wed, 14 Oct 2009 14:27:09 +0200
Subject: [Dnsmasq-discuss] Multiple resolv.conf files ...
In-Reply-To: <4AD5B000.1030708@thekelleys.org.uk>
References: <b7b14cbb0910110626x53c78e34n875c06a6a6ec00cf@mail.gmail.com>
	<2e59e6970910111207w7672041bt9c7dd932f9adb467@mail.gmail.com>
	<b7b14cbb0910140351i22e0464i2882dda9323768f@mail.gmail.com>
	<4AD5B000.1030708@thekelleys.org.uk>
Message-ID: <b7b14cbb0910140527o680656b6q5e31c161e5ccee43@mail.gmail.com>

Greetings ...

 Thanks for the quick responce ...

>>>> ?Been looking into creating a Linux Router that can be used for
>>>> internet connection fall over, when the primary internet connection
>>>> fails, I would like to have the router switch to the backup. ?I know
>>>> that DnsMasq can't really help with this, but, I'm looking at making a
>>>> second connection, using 3G ...
>>>>
>>>> ?So, my problem is that without making very complicated scripts, could
>>>> we not make DnsMasq be able to read multiple resolv.conf type files
>>>
>>> Simon already has this covered, read your man page, which says
>>> "Dnsmasq can be told to poll more than one resolv.conf file".
>>
>> ?Arh, yes ... I did see this and hoped that it would work similar to
>> other multiple options ...
>>
>>> http://www.thekelleys.org.uk/dnsmasq/docs/dnsmasq-man.html
>>
>> ?Went through that, but I did not get it to work as I had hoped ...
>>
>>> You'll solve your problem a lot faster than waiting for an answer from
>>> the list.
>>
>> ?Mmm, right and wrong ... If you read the little bit before, I was
>> hoping that I could read multiple resolv.conf type files, but testing
>> showed something that did not seem to be working as I had expected ...
>> Was hoping that either what I was epxecting, could be helped or maybe
>> another insight into another way to do a similar thing ...
>>
>>>> ?The reason I would like this, is that each internet facing interface,
>>>> using something like DHCP, would have their own DNS server list that I
>>>> could get DnsMasq to read/use, so that DNS queries would be sent down
>>>> each active internet interface ...
>>
>> ?I added to my dnsmasq.conf file two lines as follows ...
>>
>> resolv-file=/etc/resolv.conf.dnsmasq-eth0
>> resolv-file=/etc/resolv.conf.dnsmasq-3g
>>
>> ?But when I restart dnsmasq, it only list the last resolv.conf file
>> and the dns server there in ... Do I need to use the -r option from
>> the command line ( init script ) or can this option be used in the
>> dnsmasq.conf file?
>>
>> Oct 14 12:49:30 hermes dnsmasq[8652]: exiting on receipt of SIGTERM
>> Oct 14 12:49:30 hermes dnsmasq[8700]: started, version 2.45 cachesize 150
>> Oct 14 12:49:30 hermes dnsmasq[8700]: compile time options: IPv6
>> GNU-getopt no-ISC-leasefile no-DBus no-I18N TFTP
>> Oct 14 12:49:30 hermes dnsmasq[8700]: DHCP, IP range 192.168.1.65 --
>> 192.168.1.115, lease time 4h
>> Oct 14 12:49:30 hermes dnsmasq[8700]: reading /etc/resolv.conf.dnsmasq-3g
>> Oct 14 12:49:30 hermes dnsmasq[8700]: using nameserver 208.67.222.222#53
>> Oct 14 12:49:30 hermes dnsmasq[8700]: read /etc/hosts - 8 addresses
>>
>> Thanks
>> Mailed
>> LeeT
>
> Dnsmasq checks the modification times of both (all) the files, and uses the
> one that was changed last. Try this:
>
> touch /etc/resolv.conf.dnsmasq-eth0
> dig www.google.com
 Arh! Okay, when I was restarting dnsmasq, I was only seeing the
changes of the last file ...

> You should see dnsmasq read the eth0 file and start to use the nameservers
> in there.
Oct 14 13:38:38 neuropeptide dnsmasq[20760]: reading /etc/resolv.conf.dnsmasq
Oct 14 13:38:38 neuropeptide dnsmasq[20760]: using nameserver 196.212.159.193#53
Oct 14 13:38:38 neuropeptide dnsmasq[20760]: using nameserver 168.210.2.2#53
Oct 14 13:38:44 neuropeptide dnsmasq[20760]: reading
/etc/resolv.conf.dnsmasq-odns
Oct 14 13:38:44 neuropeptide dnsmasq[20760]: using nameserver 208.67.220.220#53
Oct 14 13:38:44 neuropeptide dnsmasq[20760]: using nameserver 208.67.222.222#53

Should not dnsmasq list all name servers and resolv files when
restarted or started?  Touching the resolv files works like a charm
though!!

> You need to ensure that the relevant resolv.conf file is updated when a
> network link goes live and all should be well.
 dhclient or pppd ip-up or up-down should change these files, so it
should not be a problem ...

Thanks
Mailed
LeeT


From blmink at gmail.com  Wed Oct 14 14:37:10 2009
From: blmink at gmail.com (blmink at gmail.com)
Date: Wed, 14 Oct 2009 17:37:10 +0400
Subject: [Dnsmasq-discuss] dnsmasq: failed to bind server socket for
	192.168.255.1 lo: Invalid argument
Message-ID: <84b6b9d40910140637k7b5288c3i2eafae228f40ce39@mail.gmail.com>

Hi, everybody,

i start dnsmasq as following:

/usr/sbin/dnsmasq -K -D -y -Z -b -E -z -s local -S /local/ -S
/somedomain.su/192.168.255.1#54 at lo -l /tmp/dhcp.leases -r
/tmp/resolv.conf.auto.isp
--dhcp-range=lan,192.168.255.100,192.168.255.250,255.255.255.0,12h -2
eth1 -2 eth0.1

and i get following error:

dnsmasq: failed to bind server socket for 192.168.255.1 lo: Invalid argument

Is this a bug?

The system is OpenWRT on MIPSEL.
Dnsmasq versions 2.50 and 2.51 tested.

This error message appears when I use -z option.
I've tried following:
 -z
 -z -i br-lan (which has 192.168.255.1 address)
 -z -a 192.168.255.1
 -z -a 192.168.252.2
 -z -a 192.168.253.1
all these cases lead to the above error message. I think if address
doesn't matter here, so it is a bug.

Thanks in advance,
Fedor.


From sfzhi at yahoo.com  Sat Oct 17 13:04:48 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Sat, 17 Oct 2009 14:04:48 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
Message-ID: <hbcbt2$i48$1@ger.gmane.org>

Hi,

I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.

However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.

I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.

I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.

I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).

I would appreciate any kind of help with this problem.

Thank you,
Kind regards,

Sergei.



From dnsmasq.list at michaelrack.de  Sat Oct 17 15:51:08 2009
From: dnsmasq.list at michaelrack.de (Michael Rack)
Date: Sat, 17 Oct 2009 16:51:08 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <hbcbt2$i48$1@ger.gmane.org>
References: <hbcbt2$i48$1@ger.gmane.org>
Message-ID: <4AD9D9DC.8000309@michaelrack.de>

You have to setup a bridge on your linux-router and assign every dynamic 
created ppp-interface to that.

	brctl addif br0 pppx

Configure DNSMASQ to listen on the bridges interface.

	interface=br0
	bind-interfaces

And well, you're done.

Cheers,
Michael.

Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
> Hi,
>
> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>
> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>
> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>
> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>
> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>
> I would appreciate any kind of help with this problem.
>
> Thank you,
> Kind regards,
>
> Sergei.
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss


From sfzhi at yahoo.com  Sat Oct 17 19:18:39 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Sat, 17 Oct 2009 20:18:39 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <4AD9D9DC.8000309@michaelrack.de>
References: <hbcbt2$i48$1@ger.gmane.org> <4AD9D9DC.8000309@michaelrack.de>
Message-ID: <hbd1q2$a6m$1@ger.gmane.org>

Michael Rack wrote:
> You have to setup a bridge on your linux-router and assign every dynamic 
> created ppp-interface to that.
> 
> 	brctl addif br0 pppx
> 
> Configure DNSMASQ to listen on the bridges interface.
> 
> 	interface=br0
> 	bind-interfaces
> 
> And well, you're done.
> 
> Cheers,
> Michael.

Thanks for the suggestion, Michael.

I have tried it, but unfortunately it doesn't work :(.
The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
--
Kind regards,
Sergei.


> Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
>> Hi,
>>
>> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>>
>> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>>
>> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>>
>> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>>
>> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>>
>> I would appreciate any kind of help with this problem.
>>
>> Thank you,
>> Kind regards,
>>
>> Sergei.
>>
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss



From dnsmasq.list at michaelrack.de  Sat Oct 17 22:24:05 2009
From: dnsmasq.list at michaelrack.de (Michael Rack)
Date: Sat, 17 Oct 2009 23:24:05 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <hbd1q2$a6m$1@ger.gmane.org>
References: <hbcbt2$i48$1@ger.gmane.org> <4AD9D9DC.8000309@michaelrack.de>
	<hbd1q2$a6m$1@ger.gmane.org>
Message-ID: <4ADA35F5.6080300@michaelrack.de>

correct. ppp is a point to point interface and can not assigned to a 
ethernet-bridge.

But ppp-server have to serve ip-adresses without a dhcp-server?!? you 
have to manage the dynamic ip-adresses by your ppp-server.

What server is running on your machine?

You use a dhcpc-plugin, this plugin have to relay all broadcast messages 
to dnsmasq. I think the problem is located on your dhcpc-plugin.



Am 17.10.2009 20:18, schrieb Sergei Zhirikov:
> Michael Rack wrote:
>> You have to setup a bridge on your linux-router and assign every dynamic
>> created ppp-interface to that.
>>
>> 	brctl addif br0 pppx
>>
>> Configure DNSMASQ to listen on the bridges interface.
>>
>> 	interface=br0
>> 	bind-interfaces
>>
>> And well, you're done.
>>
>> Cheers,
>> Michael.
>
> Thanks for the suggestion, Michael.
>
> I have tried it, but unfortunately it doesn't work :(.
> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
> --
> Kind regards,
> Sergei.
>
>
>> Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
>>> Hi,
>>>
>>> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>>>
>>> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>>>
>>> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>>>
>>> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>>>
>>> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>>>
>>> I would appreciate any kind of help with this problem.
>>>
>>> Thank you,
>>> Kind regards,
>>>
>>> Sergei.
>>>
>>>
>>> _______________________________________________
>>> Dnsmasq-discuss mailing list
>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss


From richardvoigt at gmail.com  Sat Oct 17 22:56:36 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sat, 17 Oct 2009 16:56:36 -0500
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <hbd1q2$a6m$1@ger.gmane.org>
References: <hbcbt2$i48$1@ger.gmane.org> <4AD9D9DC.8000309@michaelrack.de>
	<hbd1q2$a6m$1@ger.gmane.org>
Message-ID: <2e59e6970910171456y50a35127q4275ce46d9c266ee@mail.gmail.com>

On Sat, Oct 17, 2009 at 1:18 PM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
> Michael Rack wrote:
>> You have to setup a bridge on your linux-router and assign every dynamic
>> created ppp-interface to that.
>>
>> ? ? ? brctl addif br0 pppx
>>
>> Configure DNSMASQ to listen on the bridges interface.
>>
>> ? ? ? interface=br0
>> ? ? ? bind-interfaces
>>
>> And well, you're done.
>>
>> Cheers,
>> Michael.
>
> Thanks for the suggestion, Michael.
>
> I have tried it, but unfortunately it doesn't work :(.
> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).

Did you create a bridge first?

brctl addbr br0

What is the output of

brctl show


Of course, an alternate way to deal with your problem is to let
dnsmasq listen on all interfaces, and use iptables to block the
traffic on the interfaces where dhcp is not wanted.


> --
> Kind regards,
> Sergei.
>
>
>> Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
>>> Hi,
>>>
>>> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>>>
>>> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>>>
>>> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>>>
>>> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>>>
>>> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>>>
>>> I would appreciate any kind of help with this problem.
>>>
>>> Thank you,
>>> Kind regards,
>>>
>>> Sergei.
>>>
>>>
>>> _______________________________________________
>>> Dnsmasq-discuss mailing list
>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From sfzhi at yahoo.com  Sun Oct 18 16:41:31 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Sun, 18 Oct 2009 17:41:31 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <2e59e6970910171456y50a35127q4275ce46d9c266ee@mail.gmail.com>
References: <hbcbt2$i48$1@ger.gmane.org>
	<4AD9D9DC.8000309@michaelrack.de>	<hbd1q2$a6m$1@ger.gmane.org>
	<2e59e6970910171456y50a35127q4275ce46d9c266ee@mail.gmail.com>
Message-ID: <hbfcvd$e2f$1@ger.gmane.org>

richardvoigt at gmail.com wrote:
> On Sat, Oct 17, 2009 at 1:18 PM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
>> Michael Rack wrote:
>>> You have to setup a bridge on your linux-router and assign every dynamic
>>> created ppp-interface to that.
>>>
>>>       brctl addif br0 pppx
>>>
>>> Configure DNSMASQ to listen on the bridges interface.
>>>
>>>       interface=br0
>>>       bind-interfaces
>>>
>>> And well, you're done.
>>>
>>> Cheers,
>>> Michael.
>> Thanks for the suggestion, Michael.
>>
>> I have tried it, but unfortunately it doesn't work :(.
>> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
> 
> Did you create a bridge first?
> 
> brctl addbr br0

Of course, I did :)

> 
> What is the output of
> 
> brctl show
> 

# brctl show
bridge name     bridge id               STP enabled     interfaces
br0             8000.000000000000       no

I don't think it is meant to work with PPP interfaces.

> Of course, an alternate way to deal with your problem is to let
> dnsmasq listen on all interfaces, and use iptables to block the
> traffic on the interfaces where dhcp is not wanted.
> 
> 
>> --
>> Kind regards,
>> Sergei.
>>
>>
>>> Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
>>>> Hi,
>>>>
>>>> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>>>>
>>>> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>>>>
>>>> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>>>>
>>>> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>>>>
>>>> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>>>>
>>>> I would appreciate any kind of help with this problem.
>>>>
>>>> Thank you,
>>>> Kind regards,
>>>>
>>>> Sergei.
>>>>
>>>>
>>>> _______________________________________________
>>>> Dnsmasq-discuss mailing list
>>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss



From sfzhi at yahoo.com  Sun Oct 18 17:02:53 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Sun, 18 Oct 2009 18:02:53 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <4ADA35F5.6080300@michaelrack.de>
References: <hbcbt2$i48$1@ger.gmane.org>
	<4AD9D9DC.8000309@michaelrack.de>	<hbd1q2$a6m$1@ger.gmane.org>
	<4ADA35F5.6080300@michaelrack.de>
Message-ID: <hbfe7f$hjo$1@ger.gmane.org>

Michael Rack wrote:
> correct. ppp is a point to point interface and can not assigned to a 
> ethernet-bridge.
> 
> But ppp-server have to serve ip-adresses without a dhcp-server?!? you 
> have to manage the dynamic ip-adresses by your ppp-server.

The whole point of what I'm trying to do is *not* doing that. I want all IP addresses to be managed in a single place, which is dnsmasq. The reason is that it also provides DNS service. If pppd were assigning IP addresses itself, the DNS server would be unaware of that. Moreover, the auto-configuration possibilities of PPP are inferior compared to DHCP.

> What server is running on your machine?

What kind of server do you mean? If you are asking about the PPP server, it's the classic pppd from http://ppp.samba.org/.

> You use a dhcpc-plugin, this plugin have to relay all broadcast messages 
> to dnsmasq. I think the problem is located on your dhcpc-plugin.

The plugin can't relay anything. It seems you misunderstood the purpose of it. The plugin is not involved in the communication with the PPP peers in any way at all. When pppd needs an IP address to assign to a peer it asks the plugin. The plugin performes DHCP negotiation with dnsmasq while the rest of pppd is waiting (completely unaware of that fact). When the plugin gets an IP address it returns it to pppd and the configuration of the PPP link continues. After that the plugin remains idle, since its job has been done. The subsequent DHCPINFORM messages come from the peer via the PPP link, so as far as pppd is concerned, those are just normal traffic.

Kind regards,
Sergei.
 
> 
> Am 17.10.2009 20:18, schrieb Sergei Zhirikov:
>> Michael Rack wrote:
>>> You have to setup a bridge on your linux-router and assign every dynamic
>>> created ppp-interface to that.
>>>
>>> 	brctl addif br0 pppx
>>>
>>> Configure DNSMASQ to listen on the bridges interface.
>>>
>>> 	interface=br0
>>> 	bind-interfaces
>>>
>>> And well, you're done.
>>>
>>> Cheers,
>>> Michael.
>> Thanks for the suggestion, Michael.
>>
>> I have tried it, but unfortunately it doesn't work :(.
>> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
>> --
>> Kind regards,
>> Sergei.
>>
>>
>>> Am 17.10.2009 14:04, schrieb Sergei Zhirikov:
>>>> Hi,
>>>>
>>>> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
>>>>
>>>> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
>>>>
>>>> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
>>>>
>>>> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
>>>>
>>>> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
>>>>
>>>> I would appreciate any kind of help with this problem.
>>>>
>>>> Thank you,
>>>> Kind regards,
>>>>
>>>> Sergei.
>>>>
>>>>
>>>> _______________________________________________
>>>> Dnsmasq-discuss mailing list
>>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss



From kiswono at gmail.com  Mon Oct 19 04:49:34 2009
From: kiswono at gmail.com (Kiswono Prayogo)
Date: Mon, 19 Oct 2009 10:49:34 +0700
Subject: [Dnsmasq-discuss] DHCP for multiple address not in the eth0
Message-ID: <99024f5f0910182049ode932dn8b2644954e08c687@mail.gmail.com>

hi, i have linux server with eth0 only: 192.168.0.1i already set dnsmasq as
DNS server for that server
i'm using L3 switch that enable DHCP relay from any addresses.
is it possible to set dnsmasq for DHCP server for addresses 10.0.0.0/8 that
not in my eth0 subnet?

Regards,
GB
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091019/178a7470/attachment.htm 

From richardvoigt at gmail.com  Mon Oct 19 05:45:14 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 18 Oct 2009 23:45:14 -0500
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <hbfcvd$e2f$1@ger.gmane.org>
References: <hbcbt2$i48$1@ger.gmane.org> <4AD9D9DC.8000309@michaelrack.de>
	<hbd1q2$a6m$1@ger.gmane.org>
	<2e59e6970910171456y50a35127q4275ce46d9c266ee@mail.gmail.com>
	<hbfcvd$e2f$1@ger.gmane.org>
Message-ID: <2e59e6970910182145q3590c311g3f999945aad8334f@mail.gmail.com>

On Sun, Oct 18, 2009 at 10:41 AM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
> richardvoigt at gmail.com wrote:
>> On Sat, Oct 17, 2009 at 1:18 PM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
>>> Michael Rack wrote:
>>>> You have to setup a bridge on your linux-router and assign every dynamic
>>>> created ppp-interface to that.
>>>>
>>>> ? ? ? brctl addif br0 pppx
>>>>
>>>> Configure DNSMASQ to listen on the bridges interface.
>>>>
>>>> ? ? ? interface=br0
>>>> ? ? ? bind-interfaces
>>>>
>>>> And well, you're done.
>>>>
>>>> Cheers,
>>>> Michael.
>>> Thanks for the suggestion, Michael.
>>>
>>> I have tried it, but unfortunately it doesn't work :(.
>>> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
>>
>> Did you create a bridge first?
>>
>> brctl addbr br0
>
> Of course, I did :)
>
>>
>> What is the output of
>>
>> brctl show
>>
>
> # brctl show
> bridge name ? ? bridge id ? ? ? ? ? ? ? STP enabled ? ? interfaces
> br0 ? ? ? ? ? ? 8000.000000000000 ? ? ? no
>
> I don't think it is meant to work with PPP interfaces.

Right, since ppp doesn't use an ethernet physical address (MAC
address) the bridge can't learn which endpoints are through which
ports.

I'm not sure how the remote peer can generate DHCPINFORM messages when
it got its address through the PPP protocol and not through DHCP at
all.  That seems really wrong, what if you had a PPP server configured
to issue addresses from its own pool, without any DHCP involved?  Are
the DHCP messages actually coming from a host or interface beyond the
far end of the tunnel (i.e. not the ppp interface on the remote
endpoint)?


From sfzhi at yahoo.com  Mon Oct 19 21:01:02 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Mon, 19 Oct 2009 22:01:02 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
	(solved)
In-Reply-To: <hbcbt2$i48$1@ger.gmane.org>
References: <hbcbt2$i48$1@ger.gmane.org>
Message-ID: <hbighv$9co$1@ger.gmane.org>

Sergei Zhirikov wrote:
> Hi,
> 
> I'm using dnsmasq as a DHCP and DNS server on my Linux home gateway, which is also a IPSec/L2TP VPN server. For IP address allocation for the VPN PPP tunnels I use ppp-dhcpc plugin, which pretends to be a DHCP relay and requests IP addresses from dnsmasq. That part works fine.
> 
> However, when the PPP link has been established the VPN client (a Windows PC in my case, but I don't think it's relevant) send DHCPINFORM request to the broadcast address 255.255.255.255. That request reaches the Linux gateway via the PPP link, but dnsmasq does not reply, because it can see that the request is coming from an interface it is not configured to serve.
> 
> I would really like dnsmasq to reply to those DHCPINFORM requests coming in through the PPP links, but I haven't been able to configure dnsmasq for that. I can't tell it to serve the PPP interface, because the interface name is assigned dynamically (thus not known in advance). I can not tell dnsmasq to serve all interfaces with some exceptions either, because then the list of the exceptions would have to contain dynamically assigned names.
> 
> I have tried some tricks with iptables. Tried to use "-j ROUTE --iif ifname" to make the packets appear coming form another interface with a fixed name that dnsmasq is configured to serve, but dnsmasq still does not reply as if it somehow can see the original interface name.
> 
> I have tried to use a pair of VETH interfaces to route the incoming packets to one end of the virtual tunnel so that they would appear to dnsmasq coming out of the other end, but that didn't work, because I didn't manage to get the routing work the way I wanted (perhaps, because I lack necessary knowledge about advanced routing).
> 
> I would appreciate any kind of help with this problem.
> 
> Thank you,
> Kind regards,
> 
> Sergei.

Thanks to everyone who tried to help. I have managed to come up with a working solution.

In case someone wants to know, the basic idea is to create a linked pair of virtual ethernet interfaces:

    ip link add name dhcp type veth peer name pipe

(I'll omit the trivial stuff like assigning IP addresses and bringing the interfaces up.)
Then direct the DHCP traffic coming from the remote PPP endpoints into one side of the tunnel:

    iptables -t mangle -A PREROUTING ! -i dhcp -m iprange --src-range 192.168.6.16-192.168.6.254 -p udp --sport 68 --dport 67 -j ROUTE --oif pipe

And configure dnsmasq to provide DHCP service on the 'dhcp' interface.

Well, I wouldn't call this the most elegant solution, but it works for me, so it's good enough for the time being, until I come up with a better way.
--
Sergei.



From sfzhi at yahoo.com  Mon Oct 19 21:01:58 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Mon, 19 Oct 2009 22:01:58 +0200
Subject: [Dnsmasq-discuss] Answering DHCPINFORM from other interfaces
In-Reply-To: <2e59e6970910182145q3590c311g3f999945aad8334f@mail.gmail.com>
References: <hbcbt2$i48$1@ger.gmane.org>
	<4AD9D9DC.8000309@michaelrack.de>	<hbd1q2$a6m$1@ger.gmane.org>	<2e59e6970910171456y50a35127q4275ce46d9c266ee@mail.gmail.com>	<hbfcvd$e2f$1@ger.gmane.org>
	<2e59e6970910182145q3590c311g3f999945aad8334f@mail.gmail.com>
Message-ID: <hbigjn$9co$2@ger.gmane.org>

richardvoigt at gmail.com wrote:
> On Sun, Oct 18, 2009 at 10:41 AM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
>> richardvoigt at gmail.com wrote:
>>> On Sat, Oct 17, 2009 at 1:18 PM, Sergei Zhirikov <sfzhi at yahoo.com> wrote:
>>>> Michael Rack wrote:
>>>>> You have to setup a bridge on your linux-router and assign every dynamic
>>>>> created ppp-interface to that.
>>>>>
>>>>>       brctl addif br0 pppx
>>>>>
>>>>> Configure DNSMASQ to listen on the bridges interface.
>>>>>
>>>>>       interface=br0
>>>>>       bind-interfaces
>>>>>
>>>>> And well, you're done.
>>>>>
>>>>> Cheers,
>>>>> Michael.
>>>> Thanks for the suggestion, Michael.
>>>>
>>>> I have tried it, but unfortunately it doesn't work :(.
>>>> The command "brctl addif br0 ppp0" fails with "can't add ppp0 to bridge br0: Invalid argument". It looks like point-to-point interfaces can not be added to a bridge (I tried also with a TUN device, just to see if it could be added, and got the same error message).
>>> Did you create a bridge first?
>>>
>>> brctl addbr br0
>> Of course, I did :)
>>
>>> What is the output of
>>>
>>> brctl show
>>>
>> # brctl show
>> bridge name     bridge id               STP enabled     interfaces
>> br0             8000.000000000000       no
>>
>> I don't think it is meant to work with PPP interfaces.
> 
> Right, since ppp doesn't use an ethernet physical address (MAC
> address) the bridge can't learn which endpoints are through which
> ports.
> 
> I'm not sure how the remote peer can generate DHCPINFORM messages when
> it got its address through the PPP protocol and not through DHCP at
> all.  That seems really wrong, what if you had a PPP server configured
> to issue addresses from its own pool, without any DHCP involved?

As far as I understand, the whole purpose of DHCPINFORM is to allow hosts that got their IP addresses in a way other than DHCP to obtain the configuration information that may not be available in that other way. In this particular case, for example, this could be NTP server, WPAD URL, etc. I don't see anything wrong with that. Most of the DHCP options are common for the whole subnet, so it doesn't really matter if the IP address was not assigned by the DHCP server, as long as it is in the correct subnet. Although in case of PPP it would probably be more correct to talk about IP ranges rather than subnets, but the idea remains the same. By the way, DHCPINFORM can also be used by the hosts that have statically configured IP addresses.

> Are the DHCP messages actually coming from a host or interface beyond the
> far end of the tunnel (i.e. not the ppp interface on the remote
> endpoint)?

There is nothing beyond the far end of the tunnel. It's just a single host, without any further routing involved. So, the DHCP messages are coming from the remote PPP endpoint.

Anyway, thank you for trying to help. I think I have solved the problem for the time being (see my other post).
--
Sergei.



From sun at freemail.hu  Tue Oct 20 08:33:42 2009
From: sun at freemail.hu (=?ISO-8859-2?Q?Nagy_P=E9ter?=)
Date: Tue, 20 Oct 2009 09:33:42 +0200
Subject: [Dnsmasq-discuss] Home network
Message-ID: <4ADD67D6.6040106@freemail.hu>

Hi all,

I have a small network at home, and it works well locally. But now, I 
should like to look out of the world.

I got a dynamic address from my ISP, and I registered it on DynDNS.com 
(e.g. nap.homelinux.net). Is it possible that my local machines can be 
reachable from the internet, as e.g. fileserver.nap.homelinux.net?
If yes, how  shall I configure  DNSMASQ?

Thanks

Peter



From richardvoigt at gmail.com  Tue Oct 20 18:12:09 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Tue, 20 Oct 2009 12:12:09 -0500
Subject: [Dnsmasq-discuss] Home network
In-Reply-To: <4ADD67D6.6040106@freemail.hu>
References: <4ADD67D6.6040106@freemail.hu>
Message-ID: <2e59e6970910201012v1d804c9dh9f039ad422f10bdb@mail.gmail.com>

No, you have only one routable IP address, so your entire network
appears to the outside world as one computer (the router).  But it can
forward incoming connections to another computer, either all to a
single computer or based on port number.  This is outside the realm of
dns, A records don't carry any port number information at all (SRV
records do, but most software doesn't use those).

You'll be using iptables for directing incoming connections.  Good
search keywords: "+iptables +dmz"

On Tue, Oct 20, 2009 at 2:33 AM, Nagy P?ter <sun at freemail.hu> wrote:
> Hi all,
>
> I have a small network at home, and it works well locally. But now, I
> should like to look out of the world.
>
> I got a dynamic address from my ISP, and I registered it on DynDNS.com
> (e.g. nap.homelinux.net). Is it possible that my local machines can be
> reachable from the internet, as e.g. fileserver.nap.homelinux.net?
> If yes, how ?shall I configure ?DNSMASQ?
>
> Thanks
>
> Peter
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From cl at isbd.net  Tue Oct 20 20:55:12 2009
From: cl at isbd.net (Chris G)
Date: Tue, 20 Oct 2009 20:55:12 +0100
Subject: [Dnsmasq-discuss] Home network
In-Reply-To: <2e59e6970910201012v1d804c9dh9f039ad422f10bdb@mail.gmail.com>
References: <4ADD67D6.6040106@freemail.hu>
	<2e59e6970910201012v1d804c9dh9f039ad422f10bdb@mail.gmail.com>
Message-ID: <20091020195512.GE25461@chris>

On Tue, Oct 20, 2009 at 12:12:09PM -0500, richardvoigt at gmail.com wrote:
> On Tue, Oct 20, 2009 at 2:33 AM, Nagy P?ter <sun at freemail.hu> wrote:
> > Hi all,
> >
> > I have a small network at home, and it works well locally. But now, I
> > should like to look out of the world.
> >
> > I got a dynamic address from my ISP, and I registered it on DynDNS.com
> > (e.g. nap.homelinux.net). Is it possible that my local machines can be
> > reachable from the internet, as e.g. fileserver.nap.homelinux.net?
> > If yes, how ?shall I configure ?DNSMASQ?
> >
> No, you have only one routable IP address, so your entire network
> appears to the outside world as one computer (the router).  But it can
> forward incoming connections to another computer, either all to a
> single computer or based on port number.  This is outside the realm of
> dns, A records don't carry any port number information at all (SRV
> records do, but most software doesn't use those).
> 
Yes, it's usually down to the firewall *and* something called (maybe)
Port Redirection under NAT in the router.  You have to tell the router
to connect incoming requests for port XX to a particular computer on
your intranet *and* you have to open the firewall to the same thing.

Quite why routers don't combine the above two into one I have never
understood but, so far, I have never found a router that does.

> You'll be using iptables for directing incoming connections.  Good
> search keywords: "+iptables +dmz"
> 
> > Thanks
> >
> > Peter
> >
> >
> > _______________________________________________
> > Dnsmasq-discuss mailing list
> > Dnsmasq-discuss at lists.thekelleys.org.uk
> > http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> >
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss

-- 
Chris Green



From psmith at gnu.org  Wed Oct 21 22:26:06 2009
From: psmith at gnu.org (Paul Smith)
Date: Wed, 21 Oct 2009 17:26:06 -0400
Subject: [Dnsmasq-discuss] Ideas for handing out "per blade" information
Message-ID: <1256160366.2856.685.camel@homebase>

Hi all.

I have an environment with lots of blades (on the order of 150-200 or
so), all getting IP information from a central server using dnsmasq as
the DHCP server.

What I need to do is provide some information to each blade when it
registers, and (some of) this information will be specific to a given
blade (said another way, it will be different for each blade).  I was
hoping to use the DHCP response to supply this information, rather than
invent some new protocol of my own, create a daemon, etc. etc.

On the server I actually already know the MAC addresses that the blades
will register with, so I was going to use that as a key to look up the
per-blade information when the DHCP request is received.

Also, ideally the information would be obtained by running a script with
the MAC address as the argument, since these addresses can change during
the runtime of the system (someone can pull a blade and put a new one,
with a different MAC, in, for example, or they could even swap two
blades which means the information I want give them will change, since
some of it is position-dependent) and I'd prefer to grab these
dynamically rather than having some kind of timeout etc.  But, if that's
not possible I can work with it.

I've been looking through the dnsmasq man page and there are a lot of
options, but I'm not sure how to use them to do what I want.

The only mapping it appears that I can do based on MAC is dhcp-host
(maybe creating a file and giving it with dhcp-hostsfile).  But the only
way I can set options, that I can see, is using dhcp-option which is
based on network-id.  There is no way to set the network-id based on the
MAC etc. that I can see.  There doesn't seem to be any way to give
specific options for a specific client.


Soooo... anyone have any ideas for me?

Thanks!


-- 
-------------------------------------------------------------------------------
 Paul D. Smith <psmith at gnu.org>          Find some GNU make tips at:
 http://www.gnu.org                      http://make.mad-scientist.net
 "Please remain calm...I may be mad, but I am a professional." --Mad Scientist



From mythtv at maxbarry.com  Thu Oct 22 04:37:10 2009
From: mythtv at maxbarry.com (Max Barry)
Date: Thu, 22 Oct 2009 14:37:10 +1100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang ups
 after repeated access to one site - dnsmasq related?]
Message-ID: <4ADFD366.1070400@maxbarry.com>


Simon Kelley wrote:
> Hugo Heden wrote:
>> But what happens if dnsmasq then
>> forwards a ("non-local") dns-query to 127.0.0.1, i.e to itself? Could
>> that not result in some kind of infinite loop weirdness?
> 
> dnsmasq knows that it's listening on 127.0.0.1 so it's clever enough not 
> to forward requests there.

Hello! Sorry to gravedig, but I'm having this problem: dnsmasq listens 
on 127.0.0.1 and forwards requests to itself.

 From /var/log/daemon.log:

Oct 15 09:24:35 eve dnsmasq[2396]: started, version 2.47 cachesize 150
Oct 15 09:24:35 eve dnsmasq[2396]: compile time options: IPv6 GNU-getopt 
DBus I18N TFTP
Oct 15 09:24:35 eve dnsmasq[2396]: no servers found in /etc/resolv.conf, 
will retry
Oct 15 09:24:35 eve dnsmasq[2396]: read /etc/hosts - 16 addresses
Oct 15 09:24:41 eve NetworkManager: <info>  starting...
...snip...
Oct 15 09:24:49 eve NetworkManager: <info>    address 192.168.0.2
Oct 15 09:24:49 eve NetworkManager: <info>    prefix 24 (255.255.255.0)
Oct 15 09:24:49 eve NetworkManager: <info>    gateway 192.168.0.1
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '127.0.0.1'
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.134.49'
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.133.193'
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.222.222'
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.220.220'
Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '192.168.0.1'
Oct 15 09:24:49 eve NetworkManager: <info>    domain name 
'vic.bigpond.net.au'
Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
of 5 (IP Configure Commit) scheduled...
Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 4 
of 5 (IP Configure Get) complete.
Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
of 5 (IP Configure Commit) started...
Oct 15 09:24:51 eve NetworkManager: <info>  (wlan1): device state 
change: 7 -> 8
Oct 15 09:24:51 eve dnsmasq[2396]: reading /etc/resolv.conf
Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.220.220#53
Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.222.222#53
Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.133.193#53
Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.134.49#53
Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 127.0.0.1#53

Note the line above. If I manually restart dnsmasq, that line instead 
looks like "dnsmasq[3853]: ignoring nameserver 127.0.0.1 - local 
interface". But each time I turn on my computer each day, and dnsmasq 
starts automatically, it's as per the above, and it listens to the local 
interface:

Oct 15 09:25:49 eve dnsmasq[2396]: query[A] www.bom.gov.au from 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: forwarded www.bom.gov.au to 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: query[A] weather.noaa.gov from 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: forwarded weather.noaa.gov to 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: query[A] imap.gmail.com from 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: forwarded imap.gmail.com to 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1

These requests time out, making everything very slow.

$ more /etc/resolv.conf
# Generated by NetworkManager
domain vic.bigpond.net.au
search vic.bigpond.net.au
nameserver 127.0.0.1
nameserver 61.9.134.49
nameserver 61.9.133.193
# NOTE: the libc resolver may not support more than 3 nameservers.
# The nameservers listed below may not be recognized.
nameserver 208.67.222.222
nameserver 208.67.220.220

This problem only began after I started using NetworkManager on Ubuntu 
(previously I had manually configured /etc/network/interfaces).

Any help appreciated!

Max.



From richardvoigt at gmail.com  Thu Oct 22 05:14:37 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Wed, 21 Oct 2009 23:14:37 -0500
Subject: [Dnsmasq-discuss] Ideas for handing out "per blade" information
In-Reply-To: <1256160366.2856.685.camel@homebase>
References: <1256160366.2856.685.camel@homebase>
Message-ID: <2e59e6970910212114m3ee9df2cvbbea191caf0de961@mail.gmail.com>

> The only mapping it appears that I can do based on MAC is dhcp-host
> (maybe creating a file and giving it with dhcp-hostsfile). ?But the only
> way I can set options, that I can see, is using dhcp-option which is
> based on network-id. ?There is no way to set the network-id based on the
> MAC etc. that I can see. ?There doesn't seem to be any way to give
> specific options for a specific client.

Part of the dhcp-host command allows you to set one or more network-id
tags, which you can match in the dhcp-option settings.  Not dynamic as
you want, but you can definitely set options with a granularity down
to individual hosts.

>
>
> Soooo... anyone have any ideas for me?
>
> Thanks!
>
>
> --
> -------------------------------------------------------------------------------
> ?Paul D. Smith <psmith at gnu.org> ? ? ? ? ?Find some GNU make tips at:
> ?http://www.gnu.org ? ? ? ? ? ? ? ? ? ? ?http://make.mad-scientist.net
> ?"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From simon at thekelleys.org.uk  Thu Oct 22 13:06:13 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 22 Oct 2009 13:06:13 +0100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang
 ups after repeated access to one site - dnsmasq related?]
In-Reply-To: <4ADFD366.1070400@maxbarry.com>
References: <4ADFD366.1070400@maxbarry.com>
Message-ID: <4AE04AB5.7050508@thekelleys.org.uk>

Max Barry wrote:
> Simon Kelley wrote:
>> Hugo Heden wrote:
>>> But what happens if dnsmasq then
>>> forwards a ("non-local") dns-query to 127.0.0.1, i.e to itself? Could
>>> that not result in some kind of infinite loop weirdness?
>> dnsmasq knows that it's listening on 127.0.0.1 so it's clever enough not 
>> to forward requests there.
> 
> Hello! Sorry to gravedig, but I'm having this problem: dnsmasq listens 
> on 127.0.0.1 and forwards requests to itself.
> 
>  From /var/log/daemon.log:
> 
> Oct 15 09:24:35 eve dnsmasq[2396]: started, version 2.47 cachesize 150
> Oct 15 09:24:35 eve dnsmasq[2396]: compile time options: IPv6 GNU-getopt 
> DBus I18N TFTP
> Oct 15 09:24:35 eve dnsmasq[2396]: no servers found in /etc/resolv.conf, 
> will retry
> Oct 15 09:24:35 eve dnsmasq[2396]: read /etc/hosts - 16 addresses
> Oct 15 09:24:41 eve NetworkManager: <info>  starting...
> ...snip...
> Oct 15 09:24:49 eve NetworkManager: <info>    address 192.168.0.2
> Oct 15 09:24:49 eve NetworkManager: <info>    prefix 24 (255.255.255.0)
> Oct 15 09:24:49 eve NetworkManager: <info>    gateway 192.168.0.1
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '127.0.0.1'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.134.49'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.133.193'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.222.222'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.220.220'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '192.168.0.1'
> Oct 15 09:24:49 eve NetworkManager: <info>    domain name 
> 'vic.bigpond.net.au'
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
> of 5 (IP Configure Commit) scheduled...
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 4 
> of 5 (IP Configure Get) complete.
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
> of 5 (IP Configure Commit) started...
> Oct 15 09:24:51 eve NetworkManager: <info>  (wlan1): device state 
> change: 7 -> 8
> Oct 15 09:24:51 eve dnsmasq[2396]: reading /etc/resolv.conf
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.220.220#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.222.222#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.133.193#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.134.49#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 127.0.0.1#53
> 
> Note the line above. If I manually restart dnsmasq, that line instead 
> looks like "dnsmasq[3853]: ignoring nameserver 127.0.0.1 - local 
> interface". But each time I turn on my computer each day, and dnsmasq 
> starts automatically, it's as per the above, and it listens to the local 
> interface:
> 
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] www.bom.gov.au from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded www.bom.gov.au to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] weather.noaa.gov from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded weather.noaa.gov to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] imap.gmail.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded imap.gmail.com to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
> 
> These requests time out, making everything very slow.
> 
> $ more /etc/resolv.conf
> # Generated by NetworkManager
> domain vic.bigpond.net.au
> search vic.bigpond.net.au
> nameserver 127.0.0.1
> nameserver 61.9.134.49
> nameserver 61.9.133.193
> # NOTE: the libc resolver may not support more than 3 nameservers.
> # The nameservers listed below may not be recognized.
> nameserver 208.67.222.222
> nameserver 208.67.220.220
> 
> This problem only began after I started using NetworkManager on Ubuntu 
> (previously I had manually configured /etc/network/interfaces).
> 
> Any help appreciated!
> 

It would be useful to get to the bottom of this.

I guess it's something to do with the order in which things happen at 
startup. Dnsmasq doesn't treat 127.0.0.1 specially: when  it starts up 
is enumerates all the interfaces on the machine and notes those with the 
"loopback" flag set. If an address turns up later in /etc/resolv.conf 
which matches one which was owned by a loopback interface during the 
enumeration, then it is rejected.

So, if, at the time dnsmasq is started, the "lo" interface either 1) 
doesn't exist, 2) doesn't have the loopback flag set or 3) doesn't have 
address 127.0.0.1 then the address 127.0.0.1 will not be weeded out.

Could you add something like

ip addr show >/tmp/interface-state

to the /etc/init.d dnsmasq start script, reboot, and post the results here?


Cheers,

Simon.




From simon at thekelleys.org.uk  Thu Oct 22 13:13:16 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 22 Oct 2009 13:13:16 +0100
Subject: [Dnsmasq-discuss] Ideas for handing out "per blade" information
In-Reply-To: <1256160366.2856.685.camel@homebase>
References: <1256160366.2856.685.camel@homebase>
Message-ID: <4AE04C5C.1050903@thekelleys.org.uk>

Paul Smith wrote:
> Hi all.
> 
> I have an environment with lots of blades (on the order of 150-200 or
> so), all getting IP information from a central server using dnsmasq as
> the DHCP server.
> 
> What I need to do is provide some information to each blade when it
> registers, and (some of) this information will be specific to a given
> blade (said another way, it will be different for each blade).  I was
> hoping to use the DHCP response to supply this information, rather than
> invent some new protocol of my own, create a daemon, etc. etc.
> 
> On the server I actually already know the MAC addresses that the blades
> will register with, so I was going to use that as a key to look up the
> per-blade information when the DHCP request is received.
> 
> Also, ideally the information would be obtained by running a script with
> the MAC address as the argument, since these addresses can change during
> the runtime of the system (someone can pull a blade and put a new one,
> with a different MAC, in, for example, or they could even swap two
> blades which means the information I want give them will change, since
> some of it is position-dependent) and I'd prefer to grab these
> dynamically rather than having some kind of timeout etc.  But, if that's
> not possible I can work with it.
> 
> I've been looking through the dnsmasq man page and there are a lot of
> options, but I'm not sure how to use them to do what I want.
> 
> The only mapping it appears that I can do based on MAC is dhcp-host
> (maybe creating a file and giving it with dhcp-hostsfile).  But the only
> way I can set options, that I can see, is using dhcp-option which is
> based on network-id.  There is no way to set the network-id based on the
> MAC etc. that I can see.  There doesn't seem to be any way to give
> specific options for a specific client.
> 
> 
> Soooo... anyone have any ideas for me?
> 
> Thanks!
> 
> 

network-ids are sadly misnamed: they have grown into much more general 
tags now. You can certainly set a net-id tag in a dhcp-host line, and 
then use that to select the options to be sent to a particular host.

There's no way to change this information completely dynamically. You 
can however put all the dhcp-host information into a file and point at 
that file with --dhcp-hostsfile, similarly with options and 
dhcp-optsfile. Unlike a general configuration, these files _are_ re-read 
when dnsmasq gets SIGHUP.

HTH

Simon.





From jnelson+dnsmasq at jamponi.net  Thu Oct 22 15:27:48 2009
From: jnelson+dnsmasq at jamponi.net (Jon Nelson)
Date: Thu, 22 Oct 2009 09:27:48 -0500
Subject: [Dnsmasq-discuss] dnsmasq and dhcp relaying
Message-ID: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com>

I've got a setup (working) that looks like this:

workstations <-> server <-> firewall <-> internet

firewall is running dnsmasq, server is using dhcrelay
Both firewall and server have iptables-based firewalls

The setup is working fine, except

1. I don't like dhcrelay
2. I get lots of deny lines in server's firewall logs, *even though
the requests work*.

Pertinent to item 2, this is what tcpdump shows (on the 'firewall'
interface of 'server').

- a workstation will broadcast or unicast for an address
- the server receives it on eth1, iptables allows it in, dhcrelay
grabs it, and sends out a new packet on eth0 (the 'firewall' side of
'server')
- dnsmasq (on 'firewall') receives it (through iptables), generates a
response, and sends it.

Here's where it gets sticky. dhcrelay receives the request on eth1,
and sends out a request on eth0. The "Relay agent IP address" in the
payload is eth1's IP address. dnsmasq sends the reply *to this
address* (eth1), rather than the address it was received from (eth0).
The dnsmasq payload appears correct, with the "Next server IP address"
and "Relay agent IP address" properly set (eth1).

This seems to be a bug in dnsmasq - shouldn't it reply using address
that the packet was received on? The payload is correct, the UDP
destination is wrong (it seems to me).

That's the reason for the deny lines in the firewall logs, btw -
iptables sees a request go out eth0 and come back in eth0 /destined
for eth1/. Should dhcrelay be putting eth0 into the "Relay agent IP
address" field?




-- 
Jon


From simon at thekelleys.org.uk  Thu Oct 22 16:06:40 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 22 Oct 2009 16:06:40 +0100
Subject: [Dnsmasq-discuss] dnsmasq and dhcp relaying
In-Reply-To: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com>
References: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com>
Message-ID: <4AE07500.3050400@thekelleys.org.uk>

Jon Nelson wrote:
> I've got a setup (working) that looks like this:
> 
> workstations <-> server <-> firewall <-> internet
> 
> firewall is running dnsmasq, server is using dhcrelay
> Both firewall and server have iptables-based firewalls
> 
> The setup is working fine, except
> 
> 1. I don't like dhcrelay
> 2. I get lots of deny lines in server's firewall logs, *even though
> the requests work*.
> 
> Pertinent to item 2, this is what tcpdump shows (on the 'firewall'
> interface of 'server').
> 
> - a workstation will broadcast or unicast for an address
> - the server receives it on eth1, iptables allows it in, dhcrelay
> grabs it, and sends out a new packet on eth0 (the 'firewall' side of
> 'server')
> - dnsmasq (on 'firewall') receives it (through iptables), generates a
> response, and sends it.
> 
> Here's where it gets sticky. dhcrelay receives the request on eth1,
> and sends out a request on eth0. The "Relay agent IP address" in the
> payload is eth1's IP address. dnsmasq sends the reply *to this
> address* (eth1), rather than the address it was received from (eth0).
> The dnsmasq payload appears correct, with the "Next server IP address"
> and "Relay agent IP address" properly set (eth1).
> 
> This seems to be a bug in dnsmasq - shouldn't it reply using address
> that the packet was received on? The payload is correct, the UDP
> destination is wrong (it seems to me).

The current behaviour is right, according to RFC2131:

    If the 'giaddr' field in a DHCP message from a client is non-zero,
    the server sends any return messages to the 'DHCP server' port on the
    BOOTP relay agent whose address appears in 'giaddr'.

> 
> That's the reason for the deny lines in the firewall logs, btw -
> iptables sees a request go out eth0 and come back in eth0 /destined
> for eth1/. Should dhcrelay be putting eth0 into the "Relay agent IP
> address" field?

No. That would completely break things because the Relay Agent IP 
address is used to determine which subnet the client is connected to and 
therefore what IP address to allocate to it. Using the address of eth0 
would result in the client getting an address on the network shared by 
the server and the firewall rather than one on the network shared by the 
  server and workstations.



The fix (if any) here lies in modifiying the firewall rules.

If you don't like dhcrelay, you could look at 
http://thekelleys.org.uk/dhcp-helper/ as an alternative.


Cheers,

Simon.



> 
> 
> 
> 



From jnelson+dnsmasq at jamponi.net  Thu Oct 22 16:29:08 2009
From: jnelson+dnsmasq at jamponi.net (Jon Nelson)
Date: Thu, 22 Oct 2009 10:29:08 -0500
Subject: [Dnsmasq-discuss] dnsmasq and dhcp relaying
In-Reply-To: <4AE07500.3050400@thekelleys.org.uk>
References: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com> 
	<4AE07500.3050400@thekelleys.org.uk>
Message-ID: <cccedfc60910220829o629daebcj2ed9819fd9420e18@mail.gmail.com>

On Thu, Oct 22, 2009 at 10:06 AM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> Jon Nelson wrote:
>>
>> I've got a setup (working) that looks like this:
>>
>> workstations <-> server <-> firewall <-> internet
>>
>> firewall is running dnsmasq, server is using dhcrelay
>> Both firewall and server have iptables-based firewalls
>>
>> The setup is working fine, except
>>
>> 1. I don't like dhcrelay
>> 2. I get lots of deny lines in server's firewall logs, *even though
>> the requests work*.
>>
>> Pertinent to item 2, this is what tcpdump shows (on the 'firewall'
>> interface of 'server').
>>
>> - a workstation will broadcast or unicast for an address
>> - the server receives it on eth1, iptables allows it in, dhcrelay
>> grabs it, and sends out a new packet on eth0 (the 'firewall' side of
>> 'server')
>> - dnsmasq (on 'firewall') receives it (through iptables), generates a
>> response, and sends it.
>>
>> Here's where it gets sticky. dhcrelay receives the request on eth1,
>> and sends out a request on eth0. The "Relay agent IP address" in the
>> payload is eth1's IP address. dnsmasq sends the reply *to this
>> address* (eth1), rather than the address it was received from (eth0).
>> The dnsmasq payload appears correct, with the "Next server IP address"
>> and "Relay agent IP address" properly set (eth1).
>>
>> This seems to be a bug in dnsmasq - shouldn't it reply using address
>> that the packet was received on? The payload is correct, the UDP
>> destination is wrong (it seems to me).
>
> The current behaviour is right, according to RFC2131:
>
> ? If the 'giaddr' field in a DHCP message from a client is non-zero,
> ? the server sends any return messages to the 'DHCP server' port on the
> ? BOOTP relay agent whose address appears in 'giaddr'.

Is "Relay agent IP address" synonymous with 'giaddr' ? For now, I will
assume it is.

>> That's the reason for the deny lines in the firewall logs, btw -
>> iptables sees a request go out eth0 and come back in eth0 /destined
>> for eth1/. Should dhcrelay be putting eth0 into the "Relay agent IP
>> address" field?
>
> No. That would completely break things because the Relay Agent IP address is
> used to determine which subnet the client is connected to and therefore what
> IP address to allocate to it. Using the address of eth0 would result in the
> client getting an address on the network shared by the server and the
> firewall rather than one on the network shared by the ?server and
> workstations.

Ah, yes - I see.

Therefore, what I'm seeing is expected from an RFC point of view - the
address being requested is in eth1's network, so dhrelay puts eth1
into giaddr. However, the request goes out eth0.  Despite being
received on eth0, dnsmasq sends its reply to the address in the
payload, rather than the address the packet was received from - and it
is this that confuses me.  Using the normal rules of packet
forwarding, packets should go back whence they came - DHCP, however,
uses contents of the payload to determine the reply and in this case
that conflicts with the actual network layout. This seems like a
broken-ness of DHCP.

> If you don't like dhcrelay, you could look at
> http://thekelleys.org.uk/dhcp-helper/ as an alternative.

I will!

-- 
Jon


From jnelson+dnsmasq at jamponi.net  Thu Oct 22 16:39:59 2009
From: jnelson+dnsmasq at jamponi.net (Jon Nelson)
Date: Thu, 22 Oct 2009 10:39:59 -0500
Subject: [Dnsmasq-discuss] dnsmasq and dhcp relaying
In-Reply-To: <4AE07500.3050400@thekelleys.org.uk>
References: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com> 
	<4AE07500.3050400@thekelleys.org.uk>
Message-ID: <cccedfc60910220839y64d5ca08x80de1a58a7b640bb@mail.gmail.com>

On Thu, Oct 22, 2009 at 10:06 AM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> If you don't like dhcrelay, you could look at
> http://thekelleys.org.uk/dhcp-helper/ as an alternative.

I've modified my firewall rules and switched to dhcp-helper.
It appears to work just great!

Thanks!


-- 
Jon


From simon at thekelleys.org.uk  Thu Oct 22 16:47:08 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 22 Oct 2009 16:47:08 +0100
Subject: [Dnsmasq-discuss] dnsmasq and dhcp relaying
In-Reply-To: <cccedfc60910220829o629daebcj2ed9819fd9420e18@mail.gmail.com>
References: <cccedfc60910220727x6e50a7d8y7af6991866d106f2@mail.gmail.com>
	<4AE07500.3050400@thekelleys.org.uk>
	<cccedfc60910220829o629daebcj2ed9819fd9420e18@mail.gmail.com>
Message-ID: <4AE07E7C.1000708@thekelleys.org.uk>

Jon Nelson wrote:
> On Thu, Oct 22, 2009 at 10:06 AM, Simon Kelley <simon at thekelleys.org.uk> wrote:
>> Jon Nelson wrote:
>>> I've got a setup (working) that looks like this:
>>>
>>> workstations <-> server <-> firewall <-> internet
>>>
>>> firewall is running dnsmasq, server is using dhcrelay
>>> Both firewall and server have iptables-based firewalls
>>>
>>> The setup is working fine, except
>>>
>>> 1. I don't like dhcrelay
>>> 2. I get lots of deny lines in server's firewall logs, *even though
>>> the requests work*.
>>>
>>> Pertinent to item 2, this is what tcpdump shows (on the 'firewall'
>>> interface of 'server').
>>>
>>> - a workstation will broadcast or unicast for an address
>>> - the server receives it on eth1, iptables allows it in, dhcrelay
>>> grabs it, and sends out a new packet on eth0 (the 'firewall' side of
>>> 'server')
>>> - dnsmasq (on 'firewall') receives it (through iptables), generates a
>>> response, and sends it.
>>>
>>> Here's where it gets sticky. dhcrelay receives the request on eth1,
>>> and sends out a request on eth0. The "Relay agent IP address" in the
>>> payload is eth1's IP address. dnsmasq sends the reply *to this
>>> address* (eth1), rather than the address it was received from (eth0).
>>> The dnsmasq payload appears correct, with the "Next server IP address"
>>> and "Relay agent IP address" properly set (eth1).
>>>
>>> This seems to be a bug in dnsmasq - shouldn't it reply using address
>>> that the packet was received on? The payload is correct, the UDP
>>> destination is wrong (it seems to me).
>> The current behaviour is right, according to RFC2131:
>>
>>   If the 'giaddr' field in a DHCP message from a client is non-zero,
>>   the server sends any return messages to the 'DHCP server' port on the
>>   BOOTP relay agent whose address appears in 'giaddr'.
> 
> Is "Relay agent IP address" synonymous with 'giaddr' ? For now, I will
> assume it is.

It is.

> 
>>> That's the reason for the deny lines in the firewall logs, btw -
>>> iptables sees a request go out eth0 and come back in eth0 /destined
>>> for eth1/. Should dhcrelay be putting eth0 into the "Relay agent IP
>>> address" field?
>> No. That would completely break things because the Relay Agent IP address is
>> used to determine which subnet the client is connected to and therefore what
>> IP address to allocate to it. Using the address of eth0 would result in the
>> client getting an address on the network shared by the server and the
>> firewall rather than one on the network shared by the  server and
>> workstations.
> 
> Ah, yes - I see.
> 
> Therefore, what I'm seeing is expected from an RFC point of view - the
> address being requested is in eth1's network, so dhrelay puts eth1
> into giaddr. However, the request goes out eth0.  Despite being
> received on eth0, dnsmasq sends its reply to the address in the
> payload, rather than the address the packet was received from - and it
> is this that confuses me.  Using the normal rules of packet
> forwarding, packets should go back whence they came - DHCP, however,
> uses contents of the payload to determine the reply and in this case
> that conflicts with the actual network layout. This seems like a
> broken-ness of DHCP.

There is quite a lot of that. It does seem to be universal in the DHCP 
spec to ignore the source address and source port of received DHCP packets.

Cheers,

Simon.


From simon at thekelleys.org.uk  Thu Oct 22 21:06:24 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 22 Oct 2009 21:06:24 +0100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang
 ups after repeated access to one site - dnsmasq related?]
In-Reply-To: <4ADFD366.1070400@maxbarry.com>
References: <4ADFD366.1070400@maxbarry.com>
Message-ID: <4AE0BB40.6040004@thekelleys.org.uk>

Max Barry wrote:
> Simon Kelley wrote:
>> Hugo Heden wrote:
>>> But what happens if dnsmasq then
>>> forwards a ("non-local") dns-query to 127.0.0.1, i.e to itself? Could
>>> that not result in some kind of infinite loop weirdness?
>> dnsmasq knows that it's listening on 127.0.0.1 so it's clever enough not 
>> to forward requests there.
> 
> Hello! Sorry to gravedig, but I'm having this problem: dnsmasq listens 
> on 127.0.0.1 and forwards requests to itself.
> 
>  From /var/log/daemon.log:
> 
> Oct 15 09:24:35 eve dnsmasq[2396]: started, version 2.47 cachesize 150
> Oct 15 09:24:35 eve dnsmasq[2396]: compile time options: IPv6 GNU-getopt 
> DBus I18N TFTP
> Oct 15 09:24:35 eve dnsmasq[2396]: no servers found in /etc/resolv.conf, 
> will retry
> Oct 15 09:24:35 eve dnsmasq[2396]: read /etc/hosts - 16 addresses
> Oct 15 09:24:41 eve NetworkManager: <info>  starting...
> ...snip...
> Oct 15 09:24:49 eve NetworkManager: <info>    address 192.168.0.2
> Oct 15 09:24:49 eve NetworkManager: <info>    prefix 24 (255.255.255.0)
> Oct 15 09:24:49 eve NetworkManager: <info>    gateway 192.168.0.1
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '127.0.0.1'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.134.49'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.133.193'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.222.222'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.220.220'
> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '192.168.0.1'
> Oct 15 09:24:49 eve NetworkManager: <info>    domain name 
> 'vic.bigpond.net.au'
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
> of 5 (IP Configure Commit) scheduled...
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 4 
> of 5 (IP Configure Get) complete.
> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
> of 5 (IP Configure Commit) started...
> Oct 15 09:24:51 eve NetworkManager: <info>  (wlan1): device state 
> change: 7 -> 8
> Oct 15 09:24:51 eve dnsmasq[2396]: reading /etc/resolv.conf
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.220.220#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.222.222#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.133.193#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.134.49#53
> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 127.0.0.1#53
> 
> Note the line above. If I manually restart dnsmasq, that line instead 
> looks like "dnsmasq[3853]: ignoring nameserver 127.0.0.1 - local 
> interface". But each time I turn on my computer each day, and dnsmasq 
> starts automatically, it's as per the above, and it listens to the local 
> interface:
> 
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] www.bom.gov.au from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded www.bom.gov.au to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] weather.noaa.gov from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded weather.noaa.gov to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] imap.gmail.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded imap.gmail.com to 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
> 
> These requests time out, making everything very slow.
> 
> $ more /etc/resolv.conf
> # Generated by NetworkManager
> domain vic.bigpond.net.au
> search vic.bigpond.net.au
> nameserver 127.0.0.1
> nameserver 61.9.134.49
> nameserver 61.9.133.193
> # NOTE: the libc resolver may not support more than 3 nameservers.
> # The nameservers listed below may not be recognized.
> nameserver 208.67.222.222
> nameserver 208.67.220.220
> 
> This problem only began after I started using NetworkManager on Ubuntu 
> (previously I had manually configured /etc/network/interfaces).
> 
> Any help appreciated!
> 
> Max.


Could you try

http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test3.tar.gz

and let me know if that fixes things?


Cheers,

Simon.


From mythtv at maxbarry.com  Thu Oct 22 23:05:41 2009
From: mythtv at maxbarry.com (Max Barry)
Date: Fri, 23 Oct 2009 09:05:41 +1100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang ups
 after repeated access to one site - dnsmasq related?]
Message-ID: <4AE0D735.7050707@maxbarry.com>


Simon Kelly wrote:
> Max Barry wrote:
>> Simon Kelley wrote:
>>> Hugo Heden wrote:
>>>> But what happens if dnsmasq then
>>>> forwards a ("non-local") dns-query to 127.0.0.1, i.e to itself? Could
>>>> that not result in some kind of infinite loop weirdness?
>>> dnsmasq knows that it's listening on 127.0.0.1 so it's clever enough not 
>>> to forward requests there.
>> 
>> Hello! Sorry to gravedig, but I'm having this problem: dnsmasq listens 
>> on 127.0.0.1 and forwards requests to itself.
>> 
>>  From /var/log/daemon.log:
>> 
>> Oct 15 09:24:35 eve dnsmasq[2396]: started, version 2.47 cachesize 150
>> Oct 15 09:24:35 eve dnsmasq[2396]: compile time options: IPv6 GNU-getopt 
>> DBus I18N TFTP
>> Oct 15 09:24:35 eve dnsmasq[2396]: no servers found in /etc/resolv.conf, 
>> will retry
>> Oct 15 09:24:35 eve dnsmasq[2396]: read /etc/hosts - 16 addresses
>> Oct 15 09:24:41 eve NetworkManager: <info>  starting...
>> ...snip...
>> Oct 15 09:24:49 eve NetworkManager: <info>    address 192.168.0.2
>> Oct 15 09:24:49 eve NetworkManager: <info>    prefix 24 (255.255.255.0)
>> Oct 15 09:24:49 eve NetworkManager: <info>    gateway 192.168.0.1
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '127.0.0.1'
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.134.49'
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.133.193'
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.222.222'
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '208.67.220.220'
>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '192.168.0.1'
>> Oct 15 09:24:49 eve NetworkManager: <info>    domain name 
>> 'vic.bigpond.net.au'
>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
>> of 5 (IP Configure Commit) scheduled...
>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 4 
>> of 5 (IP Configure Get) complete.
>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
>> of 5 (IP Configure Commit) started...
>> Oct 15 09:24:51 eve NetworkManager: <info>  (wlan1): device state 
>> change: 7 -> 8
>> Oct 15 09:24:51 eve dnsmasq[2396]: reading /etc/resolv.conf
>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.220.220#53
>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.222.222#53
>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.133.193#53
>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.134.49#53
>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 127.0.0.1#53
>> 
>> Note the line above. If I manually restart dnsmasq, that line instead 
>> looks like "dnsmasq[3853]: ignoring nameserver 127.0.0.1 - local 
>> interface". But each time I turn on my computer each day, and dnsmasq 
>> starts automatically, it's as per the above, and it listens to the local 
>> interface:
>> 
>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] www.bom.gov.au from 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded www.bom.gov.au to 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] weather.noaa.gov from 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded weather.noaa.gov to 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] imap.gmail.com from 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded imap.gmail.com to 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 127.0.0.1
>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
>> 
>> These requests time out, making everything very slow.
>> 
>> $ more /etc/resolv.conf
>> # Generated by NetworkManager
>> domain vic.bigpond.net.au
>> search vic.bigpond.net.au
>> nameserver 127.0.0.1
>> nameserver 61.9.134.49
>> nameserver 61.9.133.193
>> # NOTE: the libc resolver may not support more than 3 nameservers.
>> # The nameservers listed below may not be recognized.
>> nameserver 208.67.222.222
>> nameserver 208.67.220.220
>> 
>> This problem only began after I started using NetworkManager on Ubuntu 
>> (previously I had manually configured /etc/network/interfaces).
>> 
>> Any help appreciated!
>> 
> 
> It would be useful to get to the bottom of this.
> 
> I guess it's something to do with the order in which things happen at 
> startup. Dnsmasq doesn't treat 127.0.0.1 specially: when  it starts up 
> is enumerates all the interfaces on the machine and notes those with the 
> "loopback" flag set. If an address turns up later in /etc/resolv.conf 
> which matches one which was owned by a loopback interface during the 
> enumeration, then it is rejected.
> 
> So, if, at the time dnsmasq is started, the "lo" interface either 1) 
> doesn't exist, 2) doesn't have the loopback flag set or 3) doesn't have 
> address 127.0.0.1 then the address 127.0.0.1 will not be weeded out.
> 
> Could you add something like
> 
> ip addr show >/tmp/interface-state
> 
> to the /etc/init.d dnsmasq start script, reboot, and post the results here?
> 
> 
> Cheers,
> 
> Simon.

Thanks for the help! Results:

1: lo: <LOOPBACK> mtu 16436 qdisc noop state DOWN
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
     link/ether 00:1f:c6:85:29:28 brd ff:ff:ff:ff:ff:ff
3: wmaster0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
     link/ieee802.11 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
4: wlan1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
     link/ether 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff

For comparison, this is what it looks like if I then restart dnsmasq 
manually:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
     inet 127.0.0.1/8 scope host lo
     inet6 ::1/128 scope host
        valid_lft forever preferred_lft forever
2: eth2: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state 
UNKNOWN qlen 1000
     link/ether 00:1f:c6:85:29:28 brd ff:ff:ff:ff:ff:ff
3: wmaster0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast 
state UNKNOWN qlen 1000
     link/ieee802.11 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
4: wlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1492 qdisc pfifo_fast 
state UP qlen 1000
     link/ether 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
     inet 192.168.0.2/24 brd 192.168.0.255 scope global wlan1
     inet6 fe80::224:1ff:fe12:878d/64 scope link
        valid_lft forever preferred_lft forever

I will see if I can get dnsmasq-2.52test3 working as per your other post.

Thanks again!

Max.



From psmith at gnu.org  Thu Oct 22 23:18:02 2009
From: psmith at gnu.org (Paul Smith)
Date: Thu, 22 Oct 2009 18:18:02 -0400
Subject: [Dnsmasq-discuss] Ideas for handing out "per blade" information
In-Reply-To: <4AE04C5C.1050903@thekelleys.org.uk>
References: <1256160366.2856.685.camel@homebase>
	<4AE04C5C.1050903@thekelleys.org.uk>
Message-ID: <1256249882.31890.82.camel@psmith-ubeta.netezza.com>

On Thu, 2009-10-22 at 13:13 +0100, Simon Kelley wrote:
> network-ids are sadly misnamed: they have grown into much more general
> tags now. You can certainly set a net-id tag in a dhcp-host line, and 
> then use that to select the options to be sent to a particular host.

I see it clearly now; I don't know how I missed it.  My only excuse is
that the summary for dhcp-options talks about <network-id> while the
summary for dhcp-hosts talks about <netid>... but that's pretty
weak! :-)

So, suppose I wanted to provide some location information in the DHCP
response packet: something like rack number + chassis number + slot
number etc.  Say a sequence of N integers.  I checked the DHCP RFCs and
didn't find a DHCP option that seemed appropriate.  I'd prefer to do
this in the most standards-compliant way possible: anyone have a
suggestion for a good option to use?

> There's no way to change this information completely dynamically. You 
> can however put all the dhcp-host information into a file and point at
> that file with --dhcp-hostsfile, similarly with options and 
> dhcp-optsfile. Unlike a general configuration, these files _are_
> re-read when dnsmasq gets SIGHUP.

I guess I need to set up a cronjob or something to do this.  I can
imagine, though, that the latency between when you put a new board in,
and when the information is available to dnsmasq to respond properly to
DHCP requests from that board, could become a source of frustration.
I'll just have to code the board to continue to do DHCP requests until
it gets back a response containing the location information I need, or
something.


Thanks for your help!



From mythtv at maxbarry.com  Fri Oct 23 03:33:57 2009
From: mythtv at maxbarry.com (Max Barry)
Date: Fri, 23 Oct 2009 13:33:57 +1100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang
 ups after repeated access to one site - dnsmasq related?]
In-Reply-To: <4AE0D735.7050707@maxbarry.com>
References: <4AE0D735.7050707@maxbarry.com>
Message-ID: <4AE11615.1090509@maxbarry.com>

On 23/10/09 09:05, Max Barry wrote:
> 
> Simon Kelly wrote:
>> Max Barry wrote:
>>> Simon Kelley wrote:
>>>> Hugo Heden wrote:
>>>>> But what happens if dnsmasq then
>>>>> forwards a ("non-local") dns-query to 127.0.0.1, i.e to itself? Could
>>>>> that not result in some kind of infinite loop weirdness?
>>>> dnsmasq knows that it's listening on 127.0.0.1 so it's clever enough 
>>>> not to forward requests there.
>>>
>>> Hello! Sorry to gravedig, but I'm having this problem: dnsmasq 
>>> listens on 127.0.0.1 and forwards requests to itself.
>>>
>>>  From /var/log/daemon.log:
>>>
>>> Oct 15 09:24:35 eve dnsmasq[2396]: started, version 2.47 cachesize 150
>>> Oct 15 09:24:35 eve dnsmasq[2396]: compile time options: IPv6 
>>> GNU-getopt DBus I18N TFTP
>>> Oct 15 09:24:35 eve dnsmasq[2396]: no servers found in 
>>> /etc/resolv.conf, will retry
>>> Oct 15 09:24:35 eve dnsmasq[2396]: read /etc/hosts - 16 addresses
>>> Oct 15 09:24:41 eve NetworkManager: <info>  starting...
>>> ...snip...
>>> Oct 15 09:24:49 eve NetworkManager: <info>    address 192.168.0.2
>>> Oct 15 09:24:49 eve NetworkManager: <info>    prefix 24 (255.255.255.0)
>>> Oct 15 09:24:49 eve NetworkManager: <info>    gateway 192.168.0.1
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '127.0.0.1'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.134.49'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '61.9.133.193'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver 
>>> '208.67.222.222'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver 
>>> '208.67.220.220'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    nameserver '192.168.0.1'
>>> Oct 15 09:24:49 eve NetworkManager: <info>    domain name 
>>> 'vic.bigpond.net.au'
>>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 
>>> 5 of 5 (IP Configure Commit) scheduled...
>>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 
>>> 4 of 5 (IP Configure Get) complete.
>>> Oct 15 09:24:49 eve NetworkManager: <info>  Activation (wlan1) Stage 
>>> 5 of 5 (IP Configure Commit) started...
>>> Oct 15 09:24:51 eve NetworkManager: <info>  (wlan1): device state 
>>> change: 7 -> 8
>>> Oct 15 09:24:51 eve dnsmasq[2396]: reading /etc/resolv.conf
>>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.220.220#53
>>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 208.67.222.222#53
>>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.133.193#53
>>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 61.9.134.49#53
>>> Oct 15 09:24:51 eve dnsmasq[2396]: using nameserver 127.0.0.1#53
>>>
>>> Note the line above. If I manually restart dnsmasq, that line instead 
>>> looks like "dnsmasq[3853]: ignoring nameserver 127.0.0.1 - local 
>>> interface". But each time I turn on my computer each day, and dnsmasq 
>>> starts automatically, it's as per the above, and it listens to the 
>>> local interface:
>>>
>>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] www.bom.gov.au from 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded www.bom.gov.au to 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] weather.noaa.gov from 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded weather.noaa.gov to 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] imap.gmail.com from 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded imap.gmail.com to 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: query[A] ntp.ubuntu.com from 
>>> 127.0.0.1
>>> Oct 15 09:25:49 eve dnsmasq[2396]: forwarded ntp.ubuntu.com to 127.0.0.1
>>>
>>> These requests time out, making everything very slow.
>>>
>>> $ more /etc/resolv.conf
>>> # Generated by NetworkManager
>>> domain vic.bigpond.net.au
>>> search vic.bigpond.net.au
>>> nameserver 127.0.0.1
>>> nameserver 61.9.134.49
>>> nameserver 61.9.133.193
>>> # NOTE: the libc resolver may not support more than 3 nameservers.
>>> # The nameservers listed below may not be recognized.
>>> nameserver 208.67.222.222
>>> nameserver 208.67.220.220
>>>
>>> This problem only began after I started using NetworkManager on 
>>> Ubuntu (previously I had manually configured /etc/network/interfaces).
>>>
>>> Any help appreciated!
>>>
>>
>> It would be useful to get to the bottom of this.
>>
>> I guess it's something to do with the order in which things happen at 
>> startup. Dnsmasq doesn't treat 127.0.0.1 specially: when  it starts up 
>> is enumerates all the interfaces on the machine and notes those with 
>> the "loopback" flag set. If an address turns up later in 
>> /etc/resolv.conf which matches one which was owned by a loopback 
>> interface during the enumeration, then it is rejected.
>>
>> So, if, at the time dnsmasq is started, the "lo" interface either 1) 
>> doesn't exist, 2) doesn't have the loopback flag set or 3) doesn't 
>> have address 127.0.0.1 then the address 127.0.0.1 will not be weeded out.
>>
>> Could you add something like
>>
>> ip addr show >/tmp/interface-state
>>
>> to the /etc/init.d dnsmasq start script, reboot, and post the results 
>> here?
>>
>>
>> Cheers,
>>
>> Simon.
> 
> Thanks for the help! Results:
> 
> 1: lo: <LOOPBACK> mtu 16436 qdisc noop state DOWN
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> 2: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
>     link/ether 00:1f:c6:85:29:28 brd ff:ff:ff:ff:ff:ff
> 3: wmaster0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
>     link/ieee802.11 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
> 4: wlan1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
>     link/ether 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
> 
> For comparison, this is what it looks like if I then restart dnsmasq 
> manually:
> 
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>     inet 127.0.0.1/8 scope host lo
>     inet6 ::1/128 scope host
>        valid_lft forever preferred_lft forever
> 2: eth2: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state 
> UNKNOWN qlen 1000
>     link/ether 00:1f:c6:85:29:28 brd ff:ff:ff:ff:ff:ff
> 3: wmaster0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast 
> state UNKNOWN qlen 1000
>     link/ieee802.11 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
> 4: wlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1492 qdisc pfifo_fast 
> state UP qlen 1000
>     link/ether 00:24:01:12:87:8d brd ff:ff:ff:ff:ff:ff
>     inet 192.168.0.2/24 brd 192.168.0.255 scope global wlan1
>     inet6 fe80::224:1ff:fe12:878d/64 scope link
>        valid_lft forever preferred_lft forever
> 
> I will see if I can get dnsmasq-2.52test3 working as per your other post.

Yay! dnsmasq-2.52test3 does indeed seem to work correctly. While the 'ip 
addr show' results are the same upon boot (lo registered as LOOPBACK but 
not with address 127.0.0.1), dnsmasq correctly refuses to listen to it:

Oct 23 13:12:34 eve dnsmasq[2404]: started, version 2.52test3 cachesize 150
Oct 23 13:12:34 eve dnsmasq[2404]: compile time options: IPv6 GNU-getopt 
no-DBus no-I18N DHCP TFTP
Oct 23 13:12:34 eve dnsmasq[2404]: no servers found in /etc/resolv.conf, 
will re
try
Oct 23 13:12:34 eve dnsmasq[2404]: read /etc/hosts - 16 addresses
Oct 23 13:12:40 eve NetworkManager: <info>  starting...
...snip...
Oct 23 13:12:57 eve NetworkManager: <info>  Activation (wlan1) Stage 5 
of 5 (IP Configure Commit) complete.
Oct 23 13:12:57 eve dnsmasq[2404]: query[SOA] local from 127.0.0.1
Oct 23 13:12:57 eve dnsmasq[2404]: query[SOA] local from 127.0.0.1
Oct 23 13:12:58 eve ntpd[2972]: ntpd exiting on signal 15
Oct 23 13:12:58 eve dnsmasq[2404]: reading /etc/resolv.conf
Oct 23 13:12:58 eve dnsmasq[2404]: using nameserver 208.67.220.220#53
Oct 23 13:12:58 eve dnsmasq[2404]: using nameserver 208.67.222.222#53
Oct 23 13:12:58 eve dnsmasq[2404]: using nameserver 61.9.133.193#53
Oct 23 13:12:58 eve dnsmasq[2404]: using nameserver 61.9.134.49#53
Oct 23 13:12:58 eve dnsmasq[2404]: ignoring nameserver 127.0.0.1 - local 
interface

I notice the compile-time options are slightly different this time 
around (no-DBus vs DBus, no I18N) but am not sure if that's relevant.

Thanks for the help!

Max.



From linuxuser09 at gmx.com  Fri Oct 23 03:46:40 2009
From: linuxuser09 at gmx.com (Linux User)
Date: Thu, 22 Oct 2009 22:46:40 -0400
Subject: [Dnsmasq-discuss] Dnsmasq is ignoring dhcp-option=option:router,
	192.168.0.1
Message-ID: <200910222246.40627.linuxuser09@gmx.com>

My primary computer serves as a DHCP/DNS server and sends out request to my 
second computer. Everything on the second computer is setup properly except I 
can't get my second computer to connect to the internet. Firewall is setup 
fine, I don't see in dmesg that it's denying my second computer access to the 
internet. Matter of fact, I don't see anything that says my second computer is 
trying to connect to the net. 

Things I can do on the second computer
ping 192.168.0.1 works
host google.com works
ping (google.com ip address) fail

When I issue the route command, I see: (The name of my DHCP/DNS server is 
netsrv.comp1.localnet.webwaredev.org)
default netsrv.comp1.lo 0.0.0.0 UG 0 0 eth0

I had set dhcp-option=option:router,192.168.0.1 and I think it should display:
default 192.168.0.1 0.0.0.0 UG 0 0 eth0

but it doesn't...

Any help I can get to help resolve this would be appreciated...

linuxuser09 at gmx.com
Thank You


From tim at theallinghams.com  Fri Oct 23 05:19:46 2009
From: tim at theallinghams.com (Tim Allingham)
Date: Fri, 23 Oct 2009 15:19:46 +1100
Subject: [Dnsmasq-discuss] Dnsmasq is ignoring dhcp-option=option:router,
 192.168.0.1
In-Reply-To: <200910222246.40627.linuxuser09@gmx.com>
References: <200910222246.40627.linuxuser09@gmx.com>
Message-ID: <1256271586.3637.13.camel@tim-laptop.hampton.247ads.com.au>

Have you set up any routing for the first PC to pass packets back and
forth to the net? I assume PC 1 is a linux box?


if so you'll generally need to enable ip_forward in the /proc filesystem

normally to do this is 
~# echo 1 > /proc/sys/net/ipv4/ip_forward

for routing, you'll also need to enable NAT or IP Masquerading to allow
the second PC to access the net.

if your not sure how to do this in iptables/ipfw or your preferred
firewall software, let me know I would be happy to help you out with
this outside of the list


as to your route, it is correct now, it is using the hostname to report
the default gateway, however the hostname is getting cut off due to the
column width in the route table report

ie 
netsrv.comp1.localnet.webwaredev.org
gets 
netsrv.comp1.lo<snip>


Cheers

Tim Allingham

-----Original Message-----
From: Linux User <linuxuser09 at gmx.com>
To: dnsmasq-discuss at lists.thekelleys.org.uk
Subject: [Dnsmasq-discuss] Dnsmasq is ignoring
dhcp-option=option:router, 192.168.0.1
Date: Thu, 22 Oct 2009 22:46:40 -0400


My primary computer serves as a DHCP/DNS server and sends out request to my 
second computer. Everything on the second computer is setup properly except I 
can't get my second computer to connect to the internet. Firewall is setup 
fine, I don't see in dmesg that it's denying my second computer access to the 
internet. Matter of fact, I don't see anything that says my second computer is 
trying to connect to the net. 

Things I can do on the second computer
ping 192.168.0.1 works
host google.com works
ping (google.com ip address) fail

When I issue the route command, I see: (The name of my DHCP/DNS server is 
netsrv.comp1.localnet.webwaredev.org)
default netsrv.comp1.lo 0.0.0.0 UG 0 0 eth0

I had set dhcp-option=option:router,192.168.0.1 and I think it should display:
default 192.168.0.1 0.0.0.0 UG 0 0 eth0

but it doesn't...

Any help I can get to help resolve this would be appreciated...

linuxuser09 at gmx.com
Thank You

_______________________________________________
Dnsmasq-discuss mailing list
Dnsmasq-discuss at lists.thekelleys.org.uk
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091023/5bffdad3/attachment.htm 

From linuxuser09 at gmx.com  Fri Oct 23 06:35:06 2009
From: linuxuser09 at gmx.com (Linux User)
Date: Fri, 23 Oct 2009 01:35:06 -0400
Subject: [Dnsmasq-discuss] Dnsmasq is ignoring dhcp-option=option:router,
	192.168.0.1
Message-ID: <200910230135.06680.linuxuser09@gmx.com>

On Friday 23 October 2009 12:19:46 am you wrote:
> Have you set up any routing for the first PC to pass packets back and
> forth to the net? I assume PC 1 is a linux box?
> 
> 
> if so you'll generally need to enable ip_forward in the /proc filesystem
> 
> normally to do this is
> ~# echo 1 > /proc/sys/net/ipv4/ip_forward
> 
> for routing, you'll also need to enable NAT or IP Masquerading to allow
> the second PC to access the net.
> 
> if your not sure how to do this in iptables/ipfw or your preferred
> firewall software, let me know I would be happy to help you out with
> this outside of the list
> 
> 
> as to your route, it is correct now, it is using the hostname to report
> the default gateway, however the hostname is getting cut off due to the
> column width in the route table report
> 
> ie
> netsrv.comp1.localnet.webwaredev.org
> gets
> netsrv.comp1.lo<snip>
> 
> 
> Cheers
> 
> Tim Allingham
> 
> -----Original Message-----
> From: Linux User <linuxuser09 at gmx.com>
> To: dnsmasq-discuss at lists.thekelleys.org.uk
> Subject: [Dnsmasq-discuss] Dnsmasq is ignoring
> dhcp-option=option:router, 192.168.0.1
> Date: Thu, 22 Oct 2009 22:46:40 -0400
> 
> 
> My primary computer serves as a DHCP/DNS server and sends out request to my
> second computer. Everything on the second computer is setup properly except
>  I can't get my second computer to connect to the internet. Firewall is
>  setup fine, I don't see in dmesg that it's denying my second computer
>  access to the internet. Matter of fact, I don't see anything that says my
>  second computer is trying to connect to the net.
> 
> Things I can do on the second computer
> ping 192.168.0.1 works
> host google.com works
> ping (google.com ip address) fail
> 
> When I issue the route command, I see: (The name of my DHCP/DNS server is
> netsrv.comp1.localnet.webwaredev.org)
> default netsrv.comp1.lo 0.0.0.0 UG 0 0 eth0
> 
> I had set dhcp-option=option:router,192.168.0.1 and I think it should
>  display: default 192.168.0.1 0.0.0.0 UG 0 0 eth0
> 
> but it doesn't...
> 
> Any help I can get to help resolve this would be appreciated...
> 
> linuxuser09 at gmx.com
> Thank You
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 

 Thank YOU Tim Allingham, it was echo 1 > /proc/sys/net/ipv4/ip_forward all 
along. My fwbuilder Firewall script was suppose to enabled that during boot. 
I'll don't know what happened and I'm investigating why not. Thanks again...


linuxuser09 at gmx.com
Thank You

P.S. Tim Allingham if you got allot of emails, it was meant to go to this 
list. Didn't realize until now


From simon at thekelleys.org.uk  Fri Oct 23 09:27:04 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Fri, 23 Oct 2009 09:27:04 +0100
Subject: [Dnsmasq-discuss] About 127.0.0.1 in resolv.conf [WAS: Web hang
 ups after repeated access to one site - dnsmasq related?]
In-Reply-To: <4AE11615.1090509@maxbarry.com>
References: <4AE0D735.7050707@maxbarry.com> <4AE11615.1090509@maxbarry.com>
Message-ID: <4AE168D8.2060208@thekelleys.org.uk>

Max Barry wrote:

> Yay! dnsmasq-2.52test3 does indeed seem to work correctly. While the 'ip 
> addr show' results are the same upon boot (lo registered as LOOPBACK but 
> not with address 127.0.0.1), dnsmasq correctly refuses to listen to it:

Excellent. I'll work on getting that fix released ASAP. The change is 
simply to re-check the interface configuration just before installing a 
new set of upstream servers.

> 
> I notice the compile-time options are slightly different this time 
> around (no-DBus vs DBus, no I18N) but am not sure if that's relevant.
> 

It doesn't affect the results of the test at all. If you want a dnsmasq 
binary with the same capabilities as standard, and assuming you're 
running Debian or Ubuntu, then you have all the bits needed to build 
packages which will have them

sudo apt-get build-dep dnsmasq

will pull in the tools needed, then

cd  dnsmasq-2.52; debian/rules binary

will build the dnsmasq and dnsmasq-base .debs.

> Thanks for the help!

No problem, thanks for you help.

Cheers,

Simon.



From ranceh at gmail.com  Fri Oct 23 16:34:28 2009
From: ranceh at gmail.com (Rance Hall)
Date: Fri, 23 Oct 2009 10:34:28 -0500
Subject: [Dnsmasq-discuss] Ideas for handing out "per blade" information
In-Reply-To: <1256249882.31890.82.camel@psmith-ubeta.netezza.com>
References: <1256160366.2856.685.camel@homebase>
	<4AE04C5C.1050903@thekelleys.org.uk>
	<1256249882.31890.82.camel@psmith-ubeta.netezza.com>
Message-ID: <845c0f80910230834i4d91e737r7bdf3399d88a8e17@mail.gmail.com>

On Thu, Oct 22, 2009 at 5:18 PM, Paul Smith <psmith at gnu.org> wrote:

> So, suppose I wanted to provide some location information in the DHCP
> response packet: something like rack number + chassis number + slot
> number etc. ?Say a sequence of N integers. ?I checked the DHCP RFCs and
> didn't find a DHCP option that seemed appropriate. ?I'd prefer to do
> this in the most standards-compliant way possible: anyone have a
> suggestion for a good option to use?
>
>> There's no way to change this information completely dynamically. You
>> can however put all the dhcp-host information into a file and point at
>> that file with --dhcp-hostsfile, similarly with options and
>> dhcp-optsfile. Unlike a general configuration, these files _are_
>> re-read when dnsmasq gets SIGHUP.
>
> I guess I need to set up a cronjob or something to do this. ?I can
> imagine, though, that the latency between when you put a new board in,
> and when the information is available to dnsmasq to respond properly to
> DHCP requests from that board, could become a source of frustration.
> I'll just have to code the board to continue to do DHCP requests until
> it gets back a response containing the location information I need, or
> something.
>
>
> Thanks for your help!

I have an idea that may help with part of this problem.  Ived used
this idea successfully with many other projects so theres no reason
why this wont work for you.

Keep in mind that this is designed to replace the cron job idea of
regular updates. This idea triggers updates only when needed.

<background> I originally found this while looking for a way to tell a
mail filter server about new email addresses changes as they occured
and not whenever a cron job got around to running.
</background>

Run a looping script that creates a named pipe, such that if the pipe
ever dies the script doesnt exit but instead creates the pipe again
(yes its an infinite loop, be careful). Run a second script that
watches that named pipe for input. The pipe-watcher script will
trigger a third script.  This third script will then update a dnsmasq
configuration file, and send dnsmasq a SIGHUP or whatever other signal
it needs to send.

The three script model here is a little cumbersome, but very modular
and its able to be used in an amazing variety of ways.  Let me know if
you are interested in pursuing something like this, and I can point
you in the right direction to make it happen.

As to your other question about making dnsmasq happy and following the
standards as closely as possible, keep reading, Im sure Simon or
someone else on this list who is much better at the dhcp protocol
level can help you come up with something that sense.


From toobyddum at gmail.com  Sun Oct 25 03:17:03 2009
From: toobyddum at gmail.com (toobyddum at gmail.com)
Date: Sun, 25 Oct 2009 11:17:03 +0800
Subject: [Dnsmasq-discuss] BUG: unable to handle kernel paging request at
	0040b4e0
Message-ID: <41416c740910242017m38a1c673p1fbfe55cf580dd80@mail.gmail.com>

hello, everyone~~~

I got a kernel panic after dnsmasq has running about 1 days, dnsmasq version
is 2.4.9, kernel is: 2.6.26.8

Any reply or suggestion is great appreciated. :)

dmesg details as follows:

BUG: unable to handle kernel paging request at 0040b4e0
IP: [<c025cf13>] d_instantiate+0xf/0x2b
*pde = 00000000
Oops: 0002 [#1]
Modules linked in: sch_ingress act_mirred cls_u32 sch_sfq sch_htb pcspkr
ppp_mppe ipt_TTL xt_hashlimit nf_nat_pptp nf_nat_proto_gre nf_conntrack_pptp
nf_conntrack_proto_gre ipt_REDIRECT ipt_set xt_mark xt_TCPMSS ip_set_iphash
ip_set xt_conntrack ipt_ACCOUNT iptable_raw arpt_mangle 8250_pnp 8250_pci
8250 serial_core nf_nat_ftp pppoe pppox nf_conntrack_ftp
nf_conntrack_netlink ppp_generic slhc xt_owner xt_TEE xt_condition
compat_xtables ipt_NETMAP arptable_filter arp_tables ipt_MASQUERADE ipt_ULOG
ipt_recent xt_multiport ipt_REJECT xt_DSCP xt_time xt_layer7 xt_iprange
ipt_ipp2p iptable_mangle iptable_filter iptable_nat nf_nat nf_conntrack_ipv4
ip_tables xt_connlimit xt_helper xt_string xt_MARK xt_comment xt_state
nf_conntrack xt_limit xt_mac xt_tcpudp x_tables nfnetlink ifb via_rhine e100
fbcon font bitblit softcursor vga16fb fb cfbcopyarea vgastate cfbimgblt
cfbfillrect tcp_bic

Pid: 17755, comm: dnsmasq Not tainted (2.6.26.8-i386 #182)
EIP: 0060:[<c025cf13>] EFLAGS: 00010286 CPU: 0
EIP is at d_instantiate+0xf/0x2b
EAX: 0040b4dc EBX: df010898 ECX: df009344 EDX: df0108d4
ESI: df009320 EDI: df010898 EBP: de1a0000 ESP: de1a1f48
 DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
Process dnsmasq (pid: 17755, ti=de1a0000 task=de7694a0 task.ti=de1a0000)
Stack: de0f61a0 c046fade 00000000 00000000 c05a6c16 0000001b 0000001b
df009320
       c046fb45 de0f61a0 00000000 f04b2fdc 0806d410 c0470534 df009320
00000000
       c0470cac 00000002 00000002 00000000 00000002 00000002 00000000
00000000
Call Trace:
 [<c046fade>] sock_attach_fd+0x4f/0x95
 [<c046fb45>] sock_map_fd+0x21/0x4a
 [<c0470534>] sys_socket+0x27/0x3a
 [<c0470cac>] sys_socketcall+0x51/0x181
 [<c0202e59>] sysenter_past_esp+0x6a/0x91
 =======================
Code: c3 56 89 d6 53 e8 fd fc ff ff 85 c0 89 c3 74 07 89 f0 e8 b8 07 00 00
89 d8 5b 5e c3 85 d2 89 d1 53 89 c3 74 15 8d 50 3c 8b 41 18 <89> 50 04 89 43
3c 8d 41 18 89 43 40 89 51 18 89 4b 08 89 d8 89
EIP: [<c025cf13>] d_instantiate+0xf/0x2b SS:ESP 0068:de1a1f48
Clocksource tsc unstable (delta = 18747400082 ns)
---[ end trace 4d6c85ead67accb1 ]---
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091025/201c3596/attachment.htm 

From richardvoigt at gmail.com  Sun Oct 25 04:46:07 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sat, 24 Oct 2009 23:46:07 -0500
Subject: [Dnsmasq-discuss] BUG: unable to handle kernel paging request
	at 0040b4e0
In-Reply-To: <41416c740910242017m38a1c673p1fbfe55cf580dd80@mail.gmail.com>
References: <41416c740910242017m38a1c673p1fbfe55cf580dd80@mail.gmail.com>
Message-ID: <2e59e6970910242146h37520ed9jeafd10546c66b9ff@mail.gmail.com>

On Sat, Oct 24, 2009 at 10:17 PM,  <toobyddum at gmail.com> wrote:
> hello, everyone~~~
>
> I got a kernel panic after dnsmasq has running about 1 days, dnsmasq version
> is 2.4.9, kernel is: 2.6.26.8

Most likely a hardware problem, possibly a kernel bug.  It's not
possible for an application such as dnsmasq, with no kernel
components, to create a kernel panic.

>
> Any reply or suggestion is great appreciated. :)
>
> dmesg details as follows:
>
> BUG: unable to handle kernel paging request at 0040b4e0
> IP: [<c025cf13>] d_instantiate+0xf/0x2b
> *pde = 00000000
> Oops: 0002 [#1]
> Modules linked in: sch_ingress act_mirred cls_u32 sch_sfq sch_htb pcspkr
> ppp_mppe ipt_TTL xt_hashlimit nf_nat_pptp nf_nat_proto_gre nf_conntrack_pptp
> nf_conntrack_proto_gre ipt_REDIRECT ipt_set xt_mark xt_TCPMSS ip_set_iphash
> ip_set xt_conntrack ipt_ACCOUNT iptable_raw arpt_mangle 8250_pnp 8250_pci
> 8250 serial_core nf_nat_ftp pppoe pppox nf_conntrack_ftp
> nf_conntrack_netlink ppp_generic slhc xt_owner xt_TEE xt_condition
> compat_xtables ipt_NETMAP arptable_filter arp_tables ipt_MASQUERADE ipt_ULOG
> ipt_recent xt_multiport ipt_REJECT xt_DSCP xt_time xt_layer7 xt_iprange
> ipt_ipp2p iptable_mangle iptable_filter iptable_nat nf_nat nf_conntrack_ipv4
> ip_tables xt_connlimit xt_helper xt_string xt_MARK xt_comment xt_state
> nf_conntrack xt_limit xt_mac xt_tcpudp x_tables nfnetlink ifb via_rhine e100
> fbcon font bitblit softcursor vga16fb fb cfbcopyarea vgastate cfbimgblt
> cfbfillrect tcp_bic
>
> Pid: 17755, comm: dnsmasq Not tainted (2.6.26.8-i386 #182)
> EIP: 0060:[<c025cf13>] EFLAGS: 00010286 CPU: 0
> EIP is at d_instantiate+0xf/0x2b
> EAX: 0040b4dc EBX: df010898 ECX: df009344 EDX: df0108d4
> ESI: df009320 EDI: df010898 EBP: de1a0000 ESP: de1a1f48
> ?DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
> Process dnsmasq (pid: 17755, ti=de1a0000 task=de7694a0 task.ti=de1a0000)
> Stack: de0f61a0 c046fade 00000000 00000000 c05a6c16 0000001b 0000001b
> df009320
> ?????? c046fb45 de0f61a0 00000000 f04b2fdc 0806d410 c0470534 df009320
> 00000000
> ?????? c0470cac 00000002 00000002 00000000 00000002 00000002 00000000
> 00000000
> Call Trace:
> ?[<c046fade>] sock_attach_fd+0x4f/0x95
> ?[<c046fb45>] sock_map_fd+0x21/0x4a
> ?[<c0470534>] sys_socket+0x27/0x3a
> ?[<c0470cac>] sys_socketcall+0x51/0x181
> ?[<c0202e59>] sysenter_past_esp+0x6a/0x91
> ?=======================
> Code: c3 56 89 d6 53 e8 fd fc ff ff 85 c0 89 c3 74 07 89 f0 e8 b8 07 00 00
> 89 d8 5b 5e c3 85 d2 89 d1 53 89 c3 74 15 8d 50 3c 8b 41 18 <89> 50 04 89 43
> 3c 8d 41 18 89 43 40 89 51 18 89 4b 08 89 d8 89
> EIP: [<c025cf13>] d_instantiate+0xf/0x2b SS:ESP 0068:de1a1f48
> Clocksource tsc unstable (delta = 18747400082 ns)
> ---[ end trace 4d6c85ead67accb1 ]---
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>
>


From knud.skrald at guldberg.info  Tue Oct 27 13:14:26 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Tue, 27 Oct 2009 14:14:26 +0100
Subject: [Dnsmasq-discuss] Logrotate again
Message-ID: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>

Hi

I am new to this list...

As <subj> says...I have a problem getting logrotate to work
with the log file from dnsmasq.
I have searched the earlierer postings

Problem:
When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to  
dnsmasq-date-xxx as usual logfiles.
A new dnsmasq log file is created with size of zero bytes but it keeps  
to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.  
dnsamsq is keep writing to this file instead of the newly created file.

Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command  
prompt makes dnsmasq to write to the "normal log file "dnsmasq"

to setup:
fedora 11
dnsmasq 2.46 from the fedora 11 packaged

snip from logrotate.conf
>>>>>>>>>>>>>>>
/var/log/dnsmasq {
     missingok
     daily
     rotate 3
     size 100k
     create 0640 nobody root
     postrotate
	kill -s USR2 `cat /var/run/dnsmasq.pid`
     endscript
}

Knud



From richardvoigt at gmail.com  Tue Oct 27 15:03:45 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Tue, 27 Oct 2009 10:03:45 -0500
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
Message-ID: <2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>

On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
> Hi
>
> I am new to this list...
>
> As <subj> says...I have a problem getting logrotate to work
> with the log file from dnsmasq.
> I have searched the earlierer postings
>
> Problem:
> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
> dnsmasq-date-xxx as usual logfiles.
> A new dnsmasq log file is created with size of zero bytes but it keeps
> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
> dnsamsq is keep writing to this file instead of the newly created file.
>
> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
> prompt makes dnsmasq to write to the "normal log file "dnsmasq"

Command prompt of what user?  Does it work as the logrotate user?

>
> to setup:
> fedora 11
> dnsmasq 2.46 from the fedora 11 packaged
>
> snip from logrotate.conf
>>>>>>>>>>>>>>>>
> /var/log/dnsmasq {
> ? ? missingok
> ? ? daily
> ? ? rotate 3
> ? ? size 100k
> ? ? create 0640 nobody root
> ? ? postrotate
> ? ? ? ?kill -s USR2 `cat /var/run/dnsmasq.pid`

I'd try the usual suspects, such as writing to syslog from the
postrotate step to make sure it's being executed and the pid is as
expected (the file might not be readable, etc).

> ? ? endscript
> }
>
> Knud
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From tom at punkave.com  Tue Oct 27 16:46:42 2009
From: tom at punkave.com (Tom Boutell)
Date: Tue, 27 Oct 2009 12:46:42 -0400
Subject: [Dnsmasq-discuss] Using dnsmasq to test coworkers' dev web sites
	within the office
Message-ID: <e52c06b30910270946p6f1d09a5x260709d17d6bb9c8@mail.gmail.com>

We're web developers. Each of us runs Apache on our individual
computer in order to do web development. Later we sync to staging and
production servers as needed.

For some time I've wanted to set things up so that we can check out
each other's live development sites within the office and when on the
VPN. Also I'd like to make it easy to access a dev site from a virtual
machine (eg Windows running IE6 in VMware...) without tweaking the
host file on the virtual machine to know where to find the site.

I recently set this up with dnsmasq. dnsmasq runs on our Intranet
server as our DHCP and DNS server. Anything in the tom.punk domain
resolves to my IP, so if you visit anyclientsite.tom.punk from within
the office you always reach my machine.

This is working, and that's great. But I'm wondering if it's possible
to do it more elegantly, using names rather than IPs in more
situations. Perhaps we could even use the DHCP client ID (although
that could lead to problems when a guest in the office innocently used
a conflicting ID, or when both the wireless and wired interfaces of
the same computer connect and are supposed to get the same IP - who
wins?).

We also use pptpd, which allows us to do the same trick when some or
all of us are connected via our VPN. (Yes I know about the security
concerns of PPTP, however if you use a truly random password it
doesn't seem to be practical to crack it. It's definitely not safe to
use more typical passwords with PPTP.)

This is what it looks like in dnsmasq.conf:

# These IPs must also appear in /etc/ppp/chap-secrets
# and below as dhcp-host entries with MAC addresses

# I replaced the actual fixed IP for me for privacy reasons, but this
is a fixed IP
address=/tom.punk/x.x.x.y
dhcp-host=my.mac.address,x.x.x.y

This is two places the IP must appear. Is there a way to improve on
that? Can we ditch the MAC address in favor of a DHCP client ID
somehow?

Also, in ppp/chap-secrets:

tom * obnoxious-password-here x.x.x.y

Perhaps I could get away with tom.punk rather than x.x.x.y here? Will
pppd resolve it?

If no one sees an obvious way to improve on this scheme, I'll probably
write a script to build these configuration files from a single
configuration file of names, mac addresses, passwords and fixed IPs
(or perhaps determine the fixed IPs on the fly). It's working great,
I'm just looking to make sure I'm not making it difficult for myself.

Thanks!

-- 
Tom Boutell
P'unk Avenue
215 755 1330
punkave.com
window.punkave.com


From knud.skrald at guldberg.info  Tue Oct 27 17:41:30 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Tue, 27 Oct 2009 18:41:30 +0100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
Message-ID: <4AE730CA.3020802@guldberg.info>

richardvoigt at gmail.com wrote:
> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>> Hi
>>
>> I am new to this list...
>>
>> As <subj> says...I have a problem getting logrotate to work
>> with the log file from dnsmasq.
>> I have searched the earlierer postings
>>
>> Problem:
>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>> dnsmasq-date-xxx as usual logfiles.
>> A new dnsmasq log file is created with size of zero bytes but it keeps
>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>> dnsamsq is keep writing to this file instead of the newly created file.
>>
>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
> 
> Command prompt of what user?  Does it work as the logrotate user?

root
I have tried and another user ...non of the logrotate aktivities are 
then allowed
I dont know which user the logrotate run under ..

> 
>> to setup:
>> fedora 11
>> dnsmasq 2.46 from the fedora 11 packaged
>>
>> snip from logrotate.conf
>> /var/log/dnsmasq {
>>     missingok
>>     daily
>>     rotate 3
>>     size 100k
>>     create 0640 nobody root
>>     postrotate
>>        kill -s USR2 `cat /var/run/dnsmasq.pid`
> 
> I'd try the usual suspects, such as writing to syslog from the
> postrotate step to make sure it's being executed and the pid is as
> expected (the file might not be readable, etc).

Hmmm....just tested from root user(again)...now it works as expected
By setting the size to something small...and deleting the log file from 
today

logrotate -f /etc/logrotate.conf

I will wait and see this night what happend



Knud


From tim at theallinghams.com  Tue Oct 27 23:51:00 2009
From: tim at theallinghams.com (Tim Allingham)
Date: Wed, 28 Oct 2009 10:51:00 +1100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <4AE730CA.3020802@guldberg.info>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
	<4AE730CA.3020802@guldberg.info>
Message-ID: <1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>

Have you tried restarting the service after the logrotate?

It sounds like the file handle is being kept open by dnsmasq, rather
then opened and closed per write.


Restarting the service should kick it back onto writing into the new
file, as the handle will be re-initiated.  You can configure this as a
post-rotate action in your logrotate conf.

Cheers

Tim

-----Original Message-----
From: knud.skrald at guldberg.info
To: dnsmasq-discuss at lists.thekelleys.org.uk
Subject: Re: [Dnsmasq-discuss] Logrotate again
Date: Tue, 27 Oct 2009 18:41:30 +0100


richardvoigt at gmail.com wrote:
> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>> Hi
>>
>> I am new to this list...
>>
>> As <subj> says...I have a problem getting logrotate to work
>> with the log file from dnsmasq.
>> I have searched the earlierer postings
>>
>> Problem:
>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>> dnsmasq-date-xxx as usual logfiles.
>> A new dnsmasq log file is created with size of zero bytes but it keeps
>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>> dnsamsq is keep writing to this file instead of the newly created file.
>>
>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
> 
> Command prompt of what user?  Does it work as the logrotate user?

root
I have tried and another user ...non of the logrotate aktivities are 
then allowed
I dont know which user the logrotate run under ..

> 
>> to setup:
>> fedora 11
>> dnsmasq 2.46 from the fedora 11 packaged
>>
>> snip from logrotate.conf
>> /var/log/dnsmasq {
>>     missingok
>>     daily
>>     rotate 3
>>     size 100k
>>     create 0640 nobody root
>>     postrotate
>>        kill -s USR2 `cat /var/run/dnsmasq.pid`
> 
> I'd try the usual suspects, such as writing to syslog from the
> postrotate step to make sure it's being executed and the pid is as
> expected (the file might not be readable, etc).

Hmmm....just tested from root user(again)...now it works as expected
By setting the size to something small...and deleting the log file from 
today

logrotate -f /etc/logrotate.conf

I will wait and see this night what happend



Knud

_______________________________________________
Dnsmasq-discuss mailing list
Dnsmasq-discuss at lists.thekelleys.org.uk
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091028/6ba4371f/attachment.htm 

From knud.skrald at guldberg.info  Wed Oct 28 06:51:16 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Wed, 28 Oct 2009 07:51:16 +0100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
	<4AE730CA.3020802@guldberg.info>
	<1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>
Message-ID: <20091028075116.14923vhsvho0yk10@webmail-srv2.servage.net>

Strange... this night it made the log rotation just as expected !
I have not changed anything...!
I will keep it under observation...

Suggestion:
Could SIG's be written to the logfile ?
It would be nice in a case like this to see if the SIG is received by DNSMASQ

Knud


Citat af Tim Allingham <tim at theallinghams.com>:

> Have you tried restarting the service after the logrotate?

Yes then DNSMASQ use the new file instead the old handle

>
> It sounds like the file handle is being kept open by dnsmasq, rather
> then opened and closed per write.
>
Yes my conclussion was the same

>
> Restarting the service should kick it back onto writing into the new
> file, as the handle will be re-initiated.  You can configure this as a
> post-rotate action in your logrotate conf.

Yes but so far I understand sigusr2 should actually close and reopen  
the log file

>
> Cheers
>
> Tim
>
> -----Original Message-----
> From: knud.skrald at guldberg.info
> To: dnsmasq-discuss at lists.thekelleys.org.uk
> Subject: Re: [Dnsmasq-discuss] Logrotate again
> Date: Tue, 27 Oct 2009 18:41:30 +0100
>
>
> richardvoigt at gmail.com wrote:
>> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>>> Hi
>>>
>>> I am new to this list...
>>>
>>> As <subj> says...I have a problem getting logrotate to work
>>> with the log file from dnsmasq.
>>> I have searched the earlierer postings
>>>
>>> Problem:
>>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>>> dnsmasq-date-xxx as usual logfiles.
>>> A new dnsmasq log file is created with size of zero bytes but it keeps
>>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>>> dnsamsq is keep writing to this file instead of the newly created file.
>>>
>>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
>>
>> Command prompt of what user?  Does it work as the logrotate user?
>
> root
> I have tried and another user ...non of the logrotate aktivities are
> then allowed
> I dont know which user the logrotate run under ..
>
>>
>>> to setup:
>>> fedora 11
>>> dnsmasq 2.46 from the fedora 11 packaged
>>>
>>> snip from logrotate.conf
>>> /var/log/dnsmasq {
>>>     missingok
>>>     daily
>>>     rotate 3
>>>     size 100k
>>>     create 0640 nobody root
>>>     postrotate
>>>        kill -s USR2 `cat /var/run/dnsmasq.pid`
>>
>> I'd try the usual suspects, such as writing to syslog from the
>> postrotate step to make sure it's being executed and the pid is as
>> expected (the file might not be readable, etc).
>
> Hmmm....just tested from root user(again)...now it works as expected
> By setting the size to something small...and deleting the log file from
> today
>
> logrotate -f /etc/logrotate.conf
>
> I will wait and see this night what happend
>
>
>
> Knud
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>




From knud.skrald at guldberg.info  Thu Oct 29 12:17:50 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Thu, 29 Oct 2009 13:17:50 +0100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <20091028075116.14923vhsvho0yk10@webmail-srv2.servage.net>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
	<4AE730CA.3020802@guldberg.info>
	<1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>
	<20091028075116.14923vhsvho0yk10@webmail-srv2.servage.net>
Message-ID: <20091029131750.16026u37kdtg2nla@webmail-srv2.servage.net>

Still working correct...

Strange !

Knud



> Strange... this night it made the log rotation just as expected !
> I have not changed anything...!
> I will keep it under observation...
>
> Suggestion:
> Could SIG's be written to the logfile ?
> It would be nice in a case like this to see if the SIG is received by DNSMASQ
>
> Knud
>
>
> Citat af Tim Allingham <tim at theallinghams.com>:
>
>> Have you tried restarting the service after the logrotate?
>
> Yes then DNSMASQ use the new file instead the old handle
>
>>
>> It sounds like the file handle is being kept open by dnsmasq, rather
>> then opened and closed per write.
>>
> Yes my conclussion was the same
>
>>
>> Restarting the service should kick it back onto writing into the new
>> file, as the handle will be re-initiated.  You can configure this as a
>> post-rotate action in your logrotate conf.
>
> Yes but so far I understand sigusr2 should actually close and reopen  
> the log file
>
>>
>> Cheers
>>
>> Tim
>>
>> -----Original Message-----
>> From: knud.skrald at guldberg.info
>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>> Subject: Re: [Dnsmasq-discuss] Logrotate again
>> Date: Tue, 27 Oct 2009 18:41:30 +0100
>>
>>
>> richardvoigt at gmail.com wrote:
>>> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>>>> Hi
>>>>
>>>> I am new to this list...
>>>>
>>>> As <subj> says...I have a problem getting logrotate to work
>>>> with the log file from dnsmasq.
>>>> I have searched the earlierer postings
>>>>
>>>> Problem:
>>>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>>>> dnsmasq-date-xxx as usual logfiles.
>>>> A new dnsmasq log file is created with size of zero bytes but it keeps
>>>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>>>> dnsamsq is keep writing to this file instead of the newly created file.
>>>>
>>>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>>>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
>>>
>>> Command prompt of what user?  Does it work as the logrotate user?
>>
>> root
>> I have tried and another user ...non of the logrotate aktivities are
>> then allowed
>> I dont know which user the logrotate run under ..
>>
>>>
>>>> to setup:
>>>> fedora 11
>>>> dnsmasq 2.46 from the fedora 11 packaged
>>>>
>>>> snip from logrotate.conf
>>>> /var/log/dnsmasq {
>>>>    missingok
>>>>    daily
>>>>    rotate 3
>>>>    size 100k
>>>>    create 0640 nobody root
>>>>    postrotate
>>>>       kill -s USR2 `cat /var/run/dnsmasq.pid`
>>>
>>> I'd try the usual suspects, such as writing to syslog from the
>>> postrotate step to make sure it's being executed and the pid is as
>>> expected (the file might not be readable, etc).
>>
>> Hmmm....just tested from root user(again)...now it works as expected
>> By setting the size to something small...and deleting the log file from
>> today
>>
>> logrotate -f /etc/logrotate.conf
>>
>> I will wait and see this night what happend
>>
>>
>>
>> Knud
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>
>
>




From tim at theallinghams.com  Fri Oct 30 00:09:07 2009
From: tim at theallinghams.com (Tim Allingham)
Date: Fri, 30 Oct 2009 11:09:07 +1100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <20091029131750.16026u37kdtg2nla@webmail-srv2.servage.net>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
	<4AE730CA.3020802@guldberg.info>
	<1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>
	<20091028075116.14923vhsvho0yk10@webmail-srv2.servage.net>
	<20091029131750.16026u37kdtg2nla@webmail-srv2.servage.net>
Message-ID: <1256861347.3700.2.camel@tim-laptop.hampton.247ads.com.au>

I have had this occur intermittently in the past with Apache, solution
was to restart apache on the postrotate, I'd still suggest doing this to
avoid future occurences


Cheers

Tim

-----Original Message-----
From: knud.skrald at guldberg.info
Cc: DNS-DISCUSS <dnsmasq-discuss at lists.thekelleys.org.uk>
Subject: Re: [Dnsmasq-discuss] Logrotate again
Date: Thu, 29 Oct 2009 13:17:50 +0100


Still working correct...

Strange !

Knud



> Strange... this night it made the log rotation just as expected !
> I have not changed anything...!
> I will keep it under observation...
>
> Suggestion:
> Could SIG's be written to the logfile ?
> It would be nice in a case like this to see if the SIG is received by DNSMASQ
>
> Knud
>
>
> Citat af Tim Allingham <tim at theallinghams.com>:
>
>> Have you tried restarting the service after the logrotate?
>
> Yes then DNSMASQ use the new file instead the old handle
>
>>
>> It sounds like the file handle is being kept open by dnsmasq, rather
>> then opened and closed per write.
>>
> Yes my conclussion was the same
>
>>
>> Restarting the service should kick it back onto writing into the new
>> file, as the handle will be re-initiated.  You can configure this as a
>> post-rotate action in your logrotate conf.
>
> Yes but so far I understand sigusr2 should actually close and reopen  
> the log file
>
>>
>> Cheers
>>
>> Tim
>>
>> -----Original Message-----
>> From: knud.skrald at guldberg.info
>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>> Subject: Re: [Dnsmasq-discuss] Logrotate again
>> Date: Tue, 27 Oct 2009 18:41:30 +0100
>>
>>
>> richardvoigt at gmail.com wrote:
>>> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>>>> Hi
>>>>
>>>> I am new to this list...
>>>>
>>>> As <subj> says...I have a problem getting logrotate to work
>>>> with the log file from dnsmasq.
>>>> I have searched the earlierer postings
>>>>
>>>> Problem:
>>>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>>>> dnsmasq-date-xxx as usual logfiles.
>>>> A new dnsmasq log file is created with size of zero bytes but it keeps
>>>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>>>> dnsamsq is keep writing to this file instead of the newly created file.
>>>>
>>>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>>>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
>>>
>>> Command prompt of what user?  Does it work as the logrotate user?
>>
>> root
>> I have tried and another user ...non of the logrotate aktivities are
>> then allowed
>> I dont know which user the logrotate run under ..
>>
>>>
>>>> to setup:
>>>> fedora 11
>>>> dnsmasq 2.46 from the fedora 11 packaged
>>>>
>>>> snip from logrotate.conf
>>>> /var/log/dnsmasq {
>>>>    missingok
>>>>    daily
>>>>    rotate 3
>>>>    size 100k
>>>>    create 0640 nobody root
>>>>    postrotate
>>>>       kill -s USR2 `cat /var/run/dnsmasq.pid`
>>>
>>> I'd try the usual suspects, such as writing to syslog from the
>>> postrotate step to make sure it's being executed and the pid is as
>>> expected (the file might not be readable, etc).
>>
>> Hmmm....just tested from root user(again)...now it works as expected
>> By setting the size to something small...and deleting the log file from
>> today
>>
>> logrotate -f /etc/logrotate.conf
>>
>> I will wait and see this night what happend
>>
>>
>>
>> Knud
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>
>
>



_______________________________________________
Dnsmasq-discuss mailing list
Dnsmasq-discuss at lists.thekelleys.org.uk
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091030/2579bebe/attachment.htm 

From hs.samix at gmail.com  Fri Oct 30 01:02:03 2009
From: hs.samix at gmail.com (H. S.)
Date: Thu, 29 Oct 2009 21:02:03 -0400
Subject: [Dnsmasq-discuss] dnsmasq problem: nokia cell phone does not accept
	dhcp offer
Message-ID: <cc73a6dc0910291802h469710a7nb51522437f936b6b@mail.gmail.com>

Hello,

I am trying to connect a nokia cell phone to my home wifi network. The
same phone has reportedly worked properly on another home wifi network
which has a Belkin wireless router. The phone is set to acquire an IP
address automatically, i.e. via dhcp.

BTW, my home wifi network is provided by a wireless card in a machine
running Debian Testing and 2.6.26-2-686 kernel. The cars is D-Link
DWL-G520 working in ap mode and has WPA encryption enabled.

The wireless is network provided by hostapd using madwifi driver.
Windows and Mac machines work properly with this wifi network (at least
no big problems noted, certainly not in Windows clients).

When I try to get the Nokia phone to connect to the wifi network, I get
these lines in my dnsmasq log (to see lines without wrapping, here is
the pastebin link for convenience - http://pastebin.com/m466c8fd2):
-----------------------------------------------------------
Oct 27 13:25:21 red hostapd: ath0: STA 11:22:33:44:55:66 IEEE 802.11:
disassociated
Oct 27 13:25:21 red hostapd: ath0: STA 11:22:33:44:55:66 IEEE 802.11:
associated
Oct 27 13:25:21 red hostapd: ath0: STA 11:22:33:44:55:66 RADIUS:
starting accounting session 4AE664FA-00000036
Oct 27 13:25:21 red hostapd: ath0: STA 11:22:33:44:55:66 WPA: pairwise
key handshake completed (WPA)
Oct 27 13:25:21 red hostapd: ath0: STA 11:22:33:44:55:66 WPA: group key
handshake completed (WPA)
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 Available DHCP
range: 192.168.5.150 -- 192.168.5.199
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 DHCPDISCOVER(ath0)
0.0.0.0 11:22:33:44:55:66
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 DHCPOFFER(ath0)
192.168.5.21 11:22:33:44:55:66
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 requested options:
12:hostname, 6:dns-server, 15:domain-name,
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 requested options:
1:netmask, 3:router, 28:broadcast, 120:sip-server
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 tags: known, ath0

Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 next server:
192.168.5.1
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  1
option: 53:message-type  02
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option: 54:server-identifier  192.168.5.1
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option: 51:lease-time  00:00:46:50
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option: 58:T1  00:00:23:28
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option: 59:T2  00:00:3d:86
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option:  1:netmask  255.255.255.0
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option: 28:broadcast  192.168.5.255
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option:  3:router  192.168.5.1
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  4
option:  6:dns-server  192.168.5.1
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  8
option: 15:domain-name  home.pvt
Oct 27 13:25:21 red dnsmasq-dhcp[11451]: 3875439214 sent size:  3
option: 12:hostname  NokiaCellPhone
---------------------------------------------------------------

Anybody know the problem might be?

If I switch off dnsmasq dhcp queries logging, i.e. if I decrease the verbosity
of the log, all I see are two lines of DHCPDISCOVER(ath0) and
DHCPOFFER(ath0) repeatedly in the log with no acceptance by the cell
phone.

It appears as though the phone is not accepting the dhcp offer.

However, if I give the phone a static IP address in its configuration,
it works properly on the wifi network. So it appears as though the
problem is dhcp related.

Hints? Suggestions?

Installed stuff:
$> dpkg -l *dnsmasq* hostap* | grep ^i
ii  dnsmasq       2.50-1  A small caching DNS proxy and DHCP/TFTP server
ii  dnsmasq-base  2.50-1  A small caching DNS proxy and DHCP/TFTP server
ii  hostapd       1:0.6.9-3  user space IEEE 802.11 AP and IEEE 802.1X/WPA/

Thanks.


From knud.skrald at guldberg.info  Fri Oct 30 06:58:58 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Fri, 30 Oct 2009 07:58:58 +0100
Subject: [Dnsmasq-discuss] Logrotate again
In-Reply-To: <1256861347.3700.2.camel@tim-laptop.hampton.247ads.com.au>
References: <20091027141426.21013w6nwunjcqwi@webmail-srv2.servage.net>
	<2e59e6970910270803u5a860ee1ke91705807cc3d730@mail.gmail.com>
	<4AE730CA.3020802@guldberg.info>
	<1256687460.3733.7.camel@tim-laptop.hampton.247ads.com.au>
	<20091028075116.14923vhsvho0yk10@webmail-srv2.servage.net>
	<20091029131750.16026u37kdtg2nla@webmail-srv2.servage.net>
	<1256861347.3700.2.camel@tim-laptop.hampton.247ads.com.au>
Message-ID: <20091030075858.167150jp40rj069e@webmail-srv2.servage.net>

Just for the book...

Still working....

Knud


Citat af Tim Allingham <tim at theallinghams.com>:

> I have had this occur intermittently in the past with Apache, solution
> was to restart apache on the postrotate, I'd still suggest doing this to
> avoid future occurences
>
>
> Cheers
>
> Tim
>
> -----Original Message-----
> From: knud.skrald at guldberg.info
> Cc: DNS-DISCUSS <dnsmasq-discuss at lists.thekelleys.org.uk>
> Subject: Re: [Dnsmasq-discuss] Logrotate again
> Date: Thu, 29 Oct 2009 13:17:50 +0100
>
>
> Still working correct...
>
> Strange !
>
> Knud
>
>
>
>> Strange... this night it made the log rotation just as expected !
>> I have not changed anything...!
>> I will keep it under observation...
>>
>> Suggestion:
>> Could SIG's be written to the logfile ?
>> It would be nice in a case like this to see if the SIG is received  
>> by DNSMASQ
>>
>> Knud
>>
>>
>> Citat af Tim Allingham <tim at theallinghams.com>:
>>
>>> Have you tried restarting the service after the logrotate?
>>
>> Yes then DNSMASQ use the new file instead the old handle
>>
>>>
>>> It sounds like the file handle is being kept open by dnsmasq, rather
>>> then opened and closed per write.
>>>
>> Yes my conclussion was the same
>>
>>>
>>> Restarting the service should kick it back onto writing into the new
>>> file, as the handle will be re-initiated.  You can configure this as a
>>> post-rotate action in your logrotate conf.
>>
>> Yes but so far I understand sigusr2 should actually close and reopen
>> the log file
>>
>>>
>>> Cheers
>>>
>>> Tim
>>>
>>> -----Original Message-----
>>> From: knud.skrald at guldberg.info
>>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>>> Subject: Re: [Dnsmasq-discuss] Logrotate again
>>> Date: Tue, 27 Oct 2009 18:41:30 +0100
>>>
>>>
>>> richardvoigt at gmail.com wrote:
>>>> On Tue, Oct 27, 2009 at 8:14 AM,  <knud.skrald at guldberg.info> wrote:
>>>>> Hi
>>>>>
>>>>> I am new to this list...
>>>>>
>>>>> As <subj> says...I have a problem getting logrotate to work
>>>>> with the log file from dnsmasq.
>>>>> I have searched the earlierer postings
>>>>>
>>>>> Problem:
>>>>> When the logfile (dnsmasq) is rotated, dnsmasq log file is renamed to
>>>>> dnsmasq-date-xxx as usual logfiles.
>>>>> A new dnsmasq log file is created with size of zero bytes but it keeps
>>>>> to be of size zero and dnsmasq-date-xxx is kepp growing in size eg.
>>>>> dnsamsq is keep writing to this file instead of the newly created file.
>>>>>
>>>>> Sending "kill -s SIGUSR2 `cat /var/run/dnsmasq.pid`" from a command
>>>>> prompt makes dnsmasq to write to the "normal log file "dnsmasq"
>>>>
>>>> Command prompt of what user?  Does it work as the logrotate user?
>>>
>>> root
>>> I have tried and another user ...non of the logrotate aktivities are
>>> then allowed
>>> I dont know which user the logrotate run under ..
>>>
>>>>
>>>>> to setup:
>>>>> fedora 11
>>>>> dnsmasq 2.46 from the fedora 11 packaged
>>>>>
>>>>> snip from logrotate.conf
>>>>> /var/log/dnsmasq {
>>>>>    missingok
>>>>>    daily
>>>>>    rotate 3
>>>>>    size 100k
>>>>>    create 0640 nobody root
>>>>>    postrotate
>>>>>       kill -s USR2 `cat /var/run/dnsmasq.pid`
>>>>
>>>> I'd try the usual suspects, such as writing to syslog from the
>>>> postrotate step to make sure it's being executed and the pid is as
>>>> expected (the file might not be readable, etc).
>>>
>>> Hmmm....just tested from root user(again)...now it works as expected
>>> By setting the size to something small...and deleting the log file from
>>> today
>>>
>>> logrotate -f /etc/logrotate.conf
>>>
>>> I will wait and see this night what happend
>>>
>>>
>>>
>>> Knud
>>>
>>> _______________________________________________
>>> Dnsmasq-discuss mailing list
>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>>
>>
>>
>>
>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>




From hs.samix at gmail.com  Fri Oct 30 21:57:07 2009
From: hs.samix at gmail.com (H.S.)
Date: Fri, 30 Oct 2009 17:57:07 -0400
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
Message-ID: <hcfnfk$326$1@ger.gmane.org>

Hello,

How do I specify dhcp-option 120 (for sip server) in my dnsmasq conf
file? I have tried:
dhcp-option:
dhcp-option=120,1,192.168.5.1

but after searching the web have not verified if this is the correct
method (there are reports of specifying it in hex as well).

Suggestions, please?

Thanks.
-- 

Please reply to this list only. I read this list on its corresponding
newsgroup on gmane.org. Replies sent to my email address are just
filtered to a folder in my mailbox and get periodically deleted without
ever having been read.



From simon at thekelleys.org.uk  Sun Nov  1 21:41:44 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Sun, 01 Nov 2009 21:41:44 +0000
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <hcfnfk$326$1@ger.gmane.org>
References: <hcfnfk$326$1@ger.gmane.org>
Message-ID: <4AEE0098.5060909@thekelleys.org.uk>

H.S. wrote:
> Hello,
> 
> How do I specify dhcp-option 120 (for sip server) in my dnsmasq conf
> file? I have tried:
> dhcp-option:
> dhcp-option=120,1,192.168.5.1
> 
> but after searching the web have not verified if this is the correct
> method (there are reports of specifying it in hex as well).
> 
> Suggestions, please?
> 
> Thanks.

Current version of dnsmasq know about the peculiar format of option 120, 
so you can just do:

dhcp-option=120,192.168.5.1

or taking advantage of the fact that dnsmasq now knows the names of most 
options

dhcp-option=option:sip,192.168.5.1

finally, the name version of the SIP option is supported too:

dhcp-option=option:sip,sipserver.mydomain.com


HTH

Simon.





From justin.mcateer at gmail.com  Mon Nov  2 16:16:01 2009
From: justin.mcateer at gmail.com (Justin McAteer)
Date: Mon, 2 Nov 2009 08:16:01 -0800
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <4AEE0098.5060909@thekelleys.org.uk>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
Message-ID: <39b4dffd0911020816o78a7330ftc43d94fe10fb736c@mail.gmail.com>

I believe many clients also support the setting of more than one SIP
server. I seem to recall testing this and having it work, but that's
been a while back.

Thank You,
Justin McAteer





On Sun, Nov 1, 2009 at 1:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> H.S. wrote:
>> Hello,
>>
>> How do I specify dhcp-option 120 (for sip server) in my dnsmasq conf
>> file? I have tried:
>> dhcp-option:
>> dhcp-option=120,1,192.168.5.1
>>
>> but after searching the web have not verified if this is the correct
>> method (there are reports of specifying it in hex as well).
>>
>> Suggestions, please?
>>
>> Thanks.
>
> Current version of dnsmasq know about the peculiar format of option 120,
> so you can just do:
>
> dhcp-option=120,192.168.5.1
>
> or taking advantage of the fact that dnsmasq now knows the names of most
> options
>
> dhcp-option=option:sip,192.168.5.1
>
> finally, the name version of the SIP option is supported too:
>
> dhcp-option=option:sip,sipserver.mydomain.com
>
>
> HTH
>
> Simon.
>
>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From cl at isbd.net  Mon Nov  2 22:30:14 2009
From: cl at isbd.net (Chris G)
Date: Mon, 2 Nov 2009 22:30:14 +0000
Subject: [Dnsmasq-discuss] Should I install resolvconf?
Message-ID: <20091102223014.GA1538@chris>

I just noticed that ubuntu says:-

    Suggested packages:
      resolvconf

when I install dnsmasq.  So, should I install resolvconf and will it
do me any good?

-- 
Chris Green



From cl at isbd.net  Mon Nov  2 23:32:18 2009
From: cl at isbd.net (Chris G)
Date: Mon, 2 Nov 2009 23:32:18 +0000
Subject: [Dnsmasq-discuss] Is it normal to need to edit /etc/hosts for
	dnsmasq
Message-ID: <20091102233217.GA6689@chris>

I have just moved my dnsmasq server from my desktop machine to a much
lower powered machine so I can turn my desktop off and save some
electricity. 

The new machine is called 'server' (original eh!) and has a static
address of 192.168.1.2.  It had a totally unmodified installation of
Ubuntu Server 9.10 before I installed dnsmasq. Thus its /etc/hosts
contained:- 

    127.0.0.1       localhost
    127.0.1.1       server server.isbd.net

    # The following lines are desirable for IPv6 capable hosts
    ::1     localhost ip6-localhost ip6-loopback
    fe00::0 ip6-localnet
    ff00::0 ip6-mcastprefix
    ff02::1 ip6-allnodes
    ff02::2 ip6-allrouters
    ff02::3 ip6-allhosts

Now this causes a problem - when asked for the address of 'server' it
returns 127.0.1.1 which isn't correct as the client machine will think
it itself is 'server'.

I notice I fixed this before by changing the 127.0.1.1 to the actual
IP address of the dnsmasq machine.  Is this normally necessary?

-- 
Chris Green



From daveesq at gmail.com  Tue Nov  3 13:42:08 2009
From: daveesq at gmail.com (David J. Haines)
Date: Tue, 3 Nov 2009 08:42:08 -0500
Subject: [Dnsmasq-discuss] SOA Records
Message-ID: <F750D22A-6DC7-4E42-885C-8C7A3EC76B04@gmail.com>

Hello all,

I was wondering whether there was any way for dnsmasq 2.47 to "fake"  
an SOA record for the local domain.  I'd like to have my LAN running  
in the ".local" TLD, but there's a Mac OS X 10.6.1 host that's not  
cooperating (due to Zeroconf inappropriately taking priority).  I  
understand that a possible solution is adding an SOA record  
for .local, and I'd like to give it a whirl.  Any thoughts?

Thanks,

David


From simon at thekelleys.org.uk  Tue Nov  3 13:59:24 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Tue, 03 Nov 2009 13:59:24 +0000
Subject: [Dnsmasq-discuss] SOA Records
In-Reply-To: <F750D22A-6DC7-4E42-885C-8C7A3EC76B04@gmail.com>
References: <F750D22A-6DC7-4E42-885C-8C7A3EC76B04@gmail.com>
Message-ID: <4AF0373C.4020707@thekelleys.org.uk>

David J. Haines wrote:
> Hello all,
> 
> I was wondering whether there was any way for dnsmasq 2.47 to "fake"  
> an SOA record for the local domain.  I'd like to have my LAN running  
> in the ".local" TLD, but there's a Mac OS X 10.6.1 host that's not  
> cooperating (due to Zeroconf inappropriately taking priority).  I  
> understand that a possible solution is adding an SOA record  
> for .local, and I'd like to give it a whirl.  Any thoughts?
> 

There isn't, but if it could be proved useful, such a facility could be 
added. The problem with Apple usurping the .local TLD has come up 
before, and the best advice seems to be "use another TLD", BTW.


Cheers,

Simon.



From stefano.bridi at gmail.com  Tue Nov  3 15:11:03 2009
From: stefano.bridi at gmail.com (Stefano Bridi)
Date: Tue, 3 Nov 2009 16:11:03 +0100
Subject: [Dnsmasq-discuss] use of "localise-queries"
Message-ID: <cb8bac10911030711y373f2fd0j46ea3d95e94205e9@mail.gmail.com>

Hi all,
I'm using "localise-queries" on dnsmasq to resolve the same name with
 multiple IP based on the interface they are coming from.
When a client in the dmz(192.168.0.0/24) ask for example the IP for the
 hostname "test" it will receive an ip like 192.168.0.10 but I want also that
a client on a vpn connection(10.15.0.0/24) that ask for the same hostname
 will receive the same private IP and in the meantime the clients on the lan
 interface (192.168.1.0/24) get the public IP of the server(21.22.23.24) "test"
I have added both the IP adesses for the hostname "test" in the hosts file
 and, as expected, when asked from the lan or from the vpn reply alternatively
 the private IP and the public one.
Is there a way to do that?


thanks
Stefano


From simon at thekelleys.org.uk  Tue Nov  3 20:37:27 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Tue, 03 Nov 2009 20:37:27 +0000
Subject: [Dnsmasq-discuss] use of "localise-queries"
In-Reply-To: <cb8bac10911030711y373f2fd0j46ea3d95e94205e9@mail.gmail.com>
References: <cb8bac10911030711y373f2fd0j46ea3d95e94205e9@mail.gmail.com>
Message-ID: <4AF09487.4000700@thekelleys.org.uk>

Stefano Bridi wrote:
> Hi all,
> I'm using "localise-queries" on dnsmasq to resolve the same name with
>  multiple IP based on the interface they are coming from.
> When a client in the dmz(192.168.0.0/24) ask for example the IP for the
>  hostname "test" it will receive an ip like 192.168.0.10 but I want also that
> a client on a vpn connection(10.15.0.0/24) that ask for the same hostname
>  will receive the same private IP and in the meantime the clients on the lan
>  interface (192.168.1.0/24) get the public IP of the server(21.22.23.24) "test"
> I have added both the IP adesses for the hostname "test" in the hosts file
>  and, as expected, when asked from the lan or from the vpn reply alternatively
>  the private IP and the public one.
> Is there a way to do that?
> 
> 
>

Localise-queries depends on matching the source of the query with
possible answers, which isn't what you want.

To get what you want will need different instances of dnsmasq running on
each interface, using the --interface and --bind-interfaces flags. It
will be messy, I'm afraid.


Cheers,

Simon.





From kiswono at gmail.com  Wed Nov  4 04:47:34 2009
From: kiswono at gmail.com (Kiswono Prayogo)
Date: Wed, 4 Nov 2009 11:47:34 +0700
Subject: [Dnsmasq-discuss] dnmasq slow for local network
Message-ID: <99024f5f0911032047y3b8820b4xe9847b3050010b6@mail.gmail.com>

Hi, i use dnsmasq for resolving my local network address,
here's my /etc/hosts file (real ip and domain name changed to fake):
## cat /etc/hosts
127.0.0.1 genesis
127.0.0.1       localhost
1.1.1.1 dns00.mysecretdomain.ac.id           dns00
1.1.1.1 dns.mysecretdomain.ac.id             dns
1.1.1.1 mail.mysecretdomain.ac.id           proxy
. . .
#20 more server hosts
. . .
#and 15000 more advertisement-blocking hosts that redirected ti 1.1.1.1

# cat /etc/dnsmasq.conf | egrep -a -v '^[[:space:]]*#' $1 | egrep -a
'[[:print:]]'
cache-size=1024
bogus-nxdomain=1.1.1.1
conf-file=/etc/kyz/hosts/block
conf-file=/etc/kyz/hosts/mail

## cat /etc/resolv.conf | egrep -a -v '^[[:space:]]*#' $1 | egrep -a
'[[:print:]]'
search mysecretdomain.ac.id
nameserver 208.67.222.222 # opendns
nameserver 208.67.220.220 # opendns
nameserver 1.1.1.4 # dns4
nameserver 1.1.1.3 # dns3
nameserver 1.1.1.2 # dns2

all my client dns request was redirected to 1.1.1.1
## cat /etc/network/if-up.d/custom
#some iptables clearing rules
. . .
real1 = 1.1.1.1
reals = 1.1.1.1/28
wifis = 10.0.0.0/8
staff = 192.168.0.0/16
iptables -A PREROUTING -t nat -p TCP -s $wifis -d ! $reals --dport 53 -j
DNAT --to $real1:53
iptables -A PREROUTING -t nat -p UDP -s $wifis -d ! $reals --dport 53 -j
DNAT --to $real1:53
iptables -A PREROUTING -t nat -p TCP -s $staff -d ! $reals --dport 53 -j
DNAT --to $real1:53
iptables -A PREROUTING -t nat -p UDP -s $staff -d ! $reals --dport 53 -j
DNAT --to $real1:53
. . .
#some default rules

and the interface of the gateway server are: eth0 = 192.168.1.1, eth1 =
1.1.1.1
all clients dns configuration was set to 192.168.1.1.
and if i open for example dns00.mysecretdomain.ac.id, it always lag for 3-25
seconds (and showing "Resolving host..." on Chromium or "Looking
up dns00.mysecretdomain.ac.id...") before it show the web.. and it does not
happened for non local web..
i wonder where is the lag problem source?

Regards,
Kiswono
GB
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091104/3718f9dd/attachment.htm 

From richardvoigt at gmail.com  Wed Nov  4 06:40:54 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Wed, 4 Nov 2009 00:40:54 -0600
Subject: [Dnsmasq-discuss] dnmasq slow for local network
In-Reply-To: <99024f5f0911032047y3b8820b4xe9847b3050010b6@mail.gmail.com>
References: <99024f5f0911032047y3b8820b4xe9847b3050010b6@mail.gmail.com>
Message-ID: <2e59e6970911032240h51228c4q8d38476bb0b52f04@mail.gmail.com>

On Tue, Nov 3, 2009 at 10:47 PM, Kiswono Prayogo <kiswono at gmail.com> wrote:
> Hi, i use dnsmasq for resolving my local network address,
> here's my /etc/hosts file (real ip and domain name changed to fake):
> ## cat /etc/hosts
> 127.0.0.1 genesis
> 127.0.0.1 ? ? ? localhost
> 1.1.1.1 dns00.mysecretdomain.ac.id ? ? ? ? ? dns00
> 1.1.1.1 dns.mysecretdomain.ac.id ? ? ? ? ? ? dns
> 1.1.1.1 mail.mysecretdomain.ac.id ? ? ? ? ? proxy
> . . .
> #20 more server hosts
> . . .
> #and 15000 more advertisement-blocking hosts that redirected ti 1.1.1.1
> # cat /etc/dnsmasq.conf | egrep -a -v '^[[:space:]]*#' $1 | egrep -a
> '[[:print:]]'
> cache-size=1024
> bogus-nxdomain=1.1.1.1
> conf-file=/etc/kyz/hosts/block
> conf-file=/etc/kyz/hosts/mail
> ## cat /etc/resolv.conf | egrep -a -v '^[[:space:]]*#' $1 | egrep -a
> '[[:print:]]'
> search?mysecretdomain.ac.id
> nameserver 208.67.222.222 # opendns
> nameserver 208.67.220.220 # opendns
> nameserver 1.1.1.4 # dns4
> nameserver 1.1.1.3 # dns3
> nameserver 1.1.1.2 # dns2
> all my client?dns request?was redirected to 1.1.1.1
> ## cat /etc/network/if-up.d/custom
> #some iptables clearing rules
> . . .
> real1 = 1.1.1.1
> reals = 1.1.1.1/28
> wifis = 10.0.0.0/8
> staff = 192.168.0.0/16
> iptables -A PREROUTING -t nat -p TCP -s $wifis -d ! $reals --dport 53 -j
> DNAT --to $real1:53
> iptables -A PREROUTING -t nat -p UDP -s $wifis -d ! $reals --dport 53 -j
> DNAT --to $real1:53
> iptables -A PREROUTING -t nat -p TCP -s $staff -d ! $reals --dport 53 -j
> DNAT --to $real1:53
> iptables -A PREROUTING -t nat -p UDP -s $staff -d ! $reals --dport 53 -j
> DNAT --to $real1:53
> . . .
> #some default rules
> and the interface of the gateway server are:?eth0 = 192.168.1.1, eth1 =
> 1.1.1.1
> all clients dns configuration was set to 192.168.1.1.
> and if i open for example?dns00.mysecretdomain.ac.id, it always lag for 3-25
> seconds (and showing "Resolving host..." on Chromium or "Looking
> up?dns00.mysecretdomain.ac.id...") before it show the web.. and it does not
> happened for non local web..
> i wonder where is the lag problem source?

I don't think that bogus-nxdomain applies to /etc/hosts listings, the
man page only says it covers replies from upstream servers.  So your
browser is probably actually trying to contact 1.1.1.1 and timing out.

You can verify this with "ping dns00.mysecretdomain.ac.id" and see if
it fails to get an address, or tries to contact 1.1.1.1


> Regards,
> Kiswono
> GB
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>
>


From gerrit at pmp.uni-hannover.de  Wed Nov  4 11:35:03 2009
From: gerrit at pmp.uni-hannover.de (Gerrit =?ISO-8859-1?Q?K=FChn?=)
Date: Wed, 4 Nov 2009 12:35:03 +0100
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts in
	its own domain
Message-ID: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>

Hi folks,

I have a bit of a problem here getting two dnsmasq instances on two
different machines to talk to each other properly:

I have one dnsmasq serving for my subnet (resolving DHCP and external
addresses), and another one serving for an intranet of private IPs inside
the subnet (sitting on a NAT-router). Now I have hosts in the subnet that
can access the intranet hosts via openvpn, so it would be nice if the
subnet dnsmasq could also resolve names for the intranet (most clients do
not know how to use split-dns, so I need one for all).

I set this up via the "server" statement:
server=/intranet.my.domain/<ip of intranet dnsmasq>

This works insofar that requests for hostname.intranet.my.domain are
forwarded to the intranet dnsmasq machine and get answered there. However,
I have a severe problem if I ask for an unknown intranet host like
unknown.intranet.my.domain. This gets forwarded to the intranet dnsmasq,
but it does not answer something like "unknown host" but forwards the
query back to the querying dnsmasq. From there on there queries go
ping-pong until I stop one server.

Does anyone here have a hint for me how to setup this correctly and how to
prevent queries for unknown hosts being sent back from the intranet
dnsmasq?


cu
  Gerrit


From simon at thekelleys.org.uk  Wed Nov  4 11:41:03 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 04 Nov 2009 11:41:03 +0000
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts
 in	its own domain
In-Reply-To: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>
References: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>
Message-ID: <4AF1684F.8060203@thekelleys.org.uk>

Gerrit K?hn wrote:
> Hi folks,
> 
> I have a bit of a problem here getting two dnsmasq instances on two
> different machines to talk to each other properly:
> 
> I have one dnsmasq serving for my subnet (resolving DHCP and external
> addresses), and another one serving for an intranet of private IPs inside
> the subnet (sitting on a NAT-router). Now I have hosts in the subnet that
> can access the intranet hosts via openvpn, so it would be nice if the
> subnet dnsmasq could also resolve names for the intranet (most clients do
> not know how to use split-dns, so I need one for all).
> 
> I set this up via the "server" statement:
> server=/intranet.my.domain/<ip of intranet dnsmasq>
> 
> This works insofar that requests for hostname.intranet.my.domain are
> forwarded to the intranet dnsmasq machine and get answered there. However,
> I have a severe problem if I ask for an unknown intranet host like
> unknown.intranet.my.domain. This gets forwarded to the intranet dnsmasq,
> but it does not answer something like "unknown host" but forwards the
> query back to the querying dnsmasq. From there on there queries go
> ping-pong until I stop one server.
> 
> Does anyone here have a hint for me how to setup this correctly and how to
> prevent queries for unknown hosts being sent back from the intranet
> dnsmasq?
> 


local=/intranet.my.domain/

on the intranet server will stop it forwarding any requests for that 
domain. I think that's what you need.


HTH

Simon.


From gerrit at pmp.uni-hannover.de  Wed Nov  4 12:18:45 2009
From: gerrit at pmp.uni-hannover.de (Gerrit =?ISO-8859-1?Q?K=FChn?=)
Date: Wed, 4 Nov 2009 13:18:45 +0100
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts
 in its own domain
In-Reply-To: <4AF1684F.8060203@thekelleys.org.uk>
References: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>
	<4AF1684F.8060203@thekelleys.org.uk>
Message-ID: <20091104131845.923ffb44.gerrit@pmp.uni-hannover.de>

On Wed, 04 Nov 2009 11:41:03 +0000 Simon Kelley <simon at thekelleys.org.uk>
wrote about Re: [Dnsmasq-discuss] dnsmasq forwards requests for unknown
hosts in its own domain:

SK> > Does anyone here have a hint for me how to setup this correctly and
SK> > how to prevent queries for unknown hosts being sent back from the
SK> > intranet dnsmasq?

SK> local=/intranet.my.domain/
SK> on the intranet server will stop it forwarding any requests for that 
SK> domain. I think that's what you need.

Yes, that's exactly what I needed, thanks a lot!


cu
  Gerrit


From bramzhou at gmail.com  Wed Nov  4 12:41:51 2009
From: bramzhou at gmail.com (Bo Zhou)
Date: Wed, 4 Nov 2009 13:41:51 +0100
Subject: [Dnsmasq-discuss] Questio on DHCP option 125
Message-ID: <8ade8a010911040441v5f1c826cw398cca193cddf01f@mail.gmail.com>

Hi,

I'm a new comer to this mail list. I noticed that 2 years ago, there were
some people talking about DHCP option 125 support in the dnsmasq. (
http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2007q3/001590.html)
I am just wondering what is the status for that. I checked dnsmasq v2.51,
seems it still doesn't support option 125. Anyone has news on that?

I do apologize that if there's maill thread already discussed similiar
question but could anybody give me the link as reference if there is. Thanks
a lot.

/Bram
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091104/c07396a0/attachment.htm 

From simon at thekelleys.org.uk  Wed Nov  4 13:24:29 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 04 Nov 2009 13:24:29 +0000
Subject: [Dnsmasq-discuss] Questio on DHCP option 125
In-Reply-To: <8ade8a010911040441v5f1c826cw398cca193cddf01f@mail.gmail.com>
References: <8ade8a010911040441v5f1c826cw398cca193cddf01f@mail.gmail.com>
Message-ID: <4AF1808D.7020604@thekelleys.org.uk>

Bo Zhou wrote:
> Hi,
>  
> I'm a new comer to this mail list. I noticed that 2 years ago, there 
> were some people talking about DHCP option 125 support in the dnsmasq. 
> (http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2007q3/001590.html) 
> I am just wondering what is the status for that. I checked dnsmasq 
> v2.51, seems it still doesn't support option 125. Anyone has news on that? 
>  
> I do apologize that if there's maill thread already discussed similiar 
> question but could anybody give me the link as reference if there is. 
> Thanks a lot.
>  

The thread in 2007 was the last time this was discussed, and you are 
correct, option 125 is still not supported. Kunal either didn't produce 
a patch, or didn't share it.

What is the application. The observation that you can hand-code anything 
  in hex still applies, and that may be a solution.


Cheers,

Simon.



From kiswono at gmail.com  Fri Nov  6 01:25:38 2009
From: kiswono at gmail.com (Kiswono Prayogo)
Date: Fri, 6 Nov 2009 08:25:38 +0700
Subject: [Dnsmasq-discuss] dnmasq slow for local network
In-Reply-To: <99024f5f0911032251w40f5cc59hc2620f9e8cad1b48@mail.gmail.com>
References: <99024f5f0911032047y3b8820b4xe9847b3050010b6@mail.gmail.com>
	<2e59e6970911032240h51228c4q8d38476bb0b52f04@mail.gmail.com>
	<99024f5f0911032251w40f5cc59hc2620f9e8cad1b48@mail.gmail.com>
Message-ID: <99024f5f0911051725s3a2305b3r3f4e0e4664599aea@mail.gmail.com>

ic, i found the source of problem by sniffing..
it's because i havent input the ipv6 version of ipv4 (in indonesia still use
ipv4)
and the browser always request for AAAA at first, dnsmasq did not reply for
the first time if there is no ipv6 available, and reply for the second time
without any result. then the browser request for A record then the page
shown.. there is about 3-20 seconds request delay between AAAA request and A
request if i haven't input the ipv6 version, such as:
::FFFF:myipv4 myhostname.secretdomain..

hahah ^^, thanks for help you've all provide..
(sorry my english skill not good enough to explain my problem clearly ^^)

Regards,
GB

On Wed, Nov 4, 2009 at 1:51 PM, Kiswono Prayogo <kiswono at gmail.com> wrote:

> no, what i mean that, using any web client i have, open a local web always
> having slower dns response than opening non local web.. but if i use the
> http://ipaddress/ or ping, it doesn't slow at all. i wonder why.. ^^
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091106/cfb3e2fa/attachment.htm 

From tom at rogie.be  Fri Nov  6 12:27:11 2009
From: tom at rogie.be (Tom Rogie)
Date: Fri, 06 Nov 2009 13:27:11 +0100
Subject: [Dnsmasq-discuss] 'transparent' dnsmasq servers for local network
Message-ID: <1257510431.3579.123.camel@Actarus.Vega>

Hello,

I'm having a VPN with 2 routers and a dedicated server (vpn-server) on
the internet.

The 2 routers on 2 different locations/ISPs are in fact Linksys'es with
DD-wrt firmware and dnsmasq is servicing DHCP and DNS with dnsmasq for
the local networks. The routers connect to a vpn-server (a dedicated
server at the hosting company) and so the 2 local networks are almost
transparent.

In the beginning I had 2 different localnets and everything was working
ok.

But now most of the computers got replaced by laptops and the users
traveling regularly over to the other sites. I'm also planning to have
users logging in directly to the vpn-server when they are on the field
(outside the 2 local networks, i.e. at home, on the field, using other
internet connections,...)

I would still like to have the network transparent but as we can't know
the position (network) of computers anymore I thought on replacing the 2
networks with a single one and having a 'distributed' network of dnsmasq
servers. Centralising the dhcp/dns on 1 site or on the dedicated server
is not really an option as the vpn is not guaranteed.

I configured the 2 routers and the vpnserver so that they forward local
dns queries to the other servers but doesn't seem to be working... 

What would be the best strategy for my setup?

Trogie



From hs.samix at gmail.com  Sat Nov  7 01:26:33 2009
From: hs.samix at gmail.com (H. S.)
Date: Fri, 6 Nov 2009 20:26:33 -0500
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <4AEE0098.5060909@thekelleys.org.uk>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
Message-ID: <cc73a6dc0911061726l73616d41vf6cb0f9857b26924@mail.gmail.com>

Sorry for the earlier try which sent the post only to Simon (don't use
gmail very often, usually gmane newsgroup, and I am not familiar with
the behaviour of Reply option in gmail).


On Sun, Nov 1, 2009 at 4:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> H.S. wrote:
>>
>> Hello,
>>
>> How do I specify dhcp-option 120 (for sip server) in my dnsmasq conf
>> file? I have tried:
>> dhcp-option:
>> dhcp-option=120,1,192.168.5.1
>>
>> but after searching the web have not verified if this is the correct
>> method (there are reports of specifying it in hex as well).
>>
>> Suggestions, please?
>>
>> Thanks.
>
> Current version of dnsmasq know about the peculiar format of option 120, so
> you can just do:
>
> dhcp-option=120,192.168.5.1
>
> or taking advantage of the fact that dnsmasq now knows the names of most
> options
>
> dhcp-option=option:sip,192.168.5.1
>
> finally, the name version of the SIP option is supported too:
>
> dhcp-option=option:sip,sipserver.mydomain.com
>

Thanks for this info, Simon. I will give this a shot and see how it goes.

Regards,
->HS


From hs.samix at gmail.com  Sat Nov  7 01:30:06 2009
From: hs.samix at gmail.com (H. S.)
Date: Fri, 6 Nov 2009 20:30:06 -0500
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <4AEE0098.5060909@thekelleys.org.uk>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
Message-ID: <cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>

On Sun, Nov 1, 2009 at 4:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:

>
> Current version of dnsmasq know about the peculiar format of option 120, so
> you can just do:
>
> dhcp-option=120,192.168.5.1

dnsmasq had no problem with this. Yet to try with the Nokia phone.


>
> or taking advantage of the fact that dnsmasq now knows the names of most
> options
>
> dhcp-option=option:sip,192.168.5.1

dnsmasq barfed at this:
> sudo /etc/init.d/dnsmasq restart
Restarting DNS forwarder and DHCP server: dnsmasq
dnsmasq: bad dhcp-option at line 99 of /etc/dnsmasq.d/01-home-hosts.conf
 failed!


>
> finally, the name version of the SIP option is supported too:
>
> dhcp-option=option:sip,sipserver.mydomain.com

didn't try this.

BTW, does the IP address need to be a valid SIP server for the option
to be acceptable to dnsmasq?

Thanks.


From hs.samix at gmail.com  Sat Nov  7 01:40:38 2009
From: hs.samix at gmail.com (H. S.)
Date: Fri, 6 Nov 2009 20:40:38 -0500
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
	<cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>
Message-ID: <cc73a6dc0911061740s1005632j5d745fa51b19d115@mail.gmail.com>

On Fri, Nov 6, 2009 at 8:30 PM, H. S. <hs.samix at gmail.com> wrote:
> On Sun, Nov 1, 2009 at 4:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
>
>>
>> Current version of dnsmasq know about the peculiar format of option 120, so
>> you can just do:
>>
>> dhcp-option=120,192.168.5.1
>
> dnsmasq had no problem with this. Yet to try with the Nokia phone.

Despite giving that option in dnsmasq and restarting it, the phone
still did not grab a DHCP offer :(


From cl at isbd.net  Sun Nov  8 12:08:04 2009
From: cl at isbd.net (Chris G)
Date: Sun, 8 Nov 2009 12:08:04 +0000
Subject: [Dnsmasq-discuss] dnsmasq returns addresses for non-existent hosts
	- what have I	mis-configured?
Message-ID: <20091108120803.GA14910@chris>

I have dnsmasq working quite happily on a Ubuntu Server 9.10 system
providing dns for my small SoHo network.

I have just noticed however that if I ask for the address of a
non-existent name dnsmasq returns the name of one of my hosting
service's machines.  It always returns the same address for any
non-existent name, e.g.:-

    chris$ host abcde.isbd.net
    abcde.isbd.net has address 195.74.61.93
    abcde.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
    chris$ host xyz
    xyz.isbd.net has address 195.74.61.93
    xyz.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
    chris$ host xyz.isbd.net
    xyz.isbd.net has address 195.74.61.93
    xyz.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.


I have isbd.net and isbd.co.uk hosted at Gradwell so I do have a
connection with them.  In fact things are becoming clearer now,
195.74.61.93 is the (quite correct) address returned when you
look up isbd.net.

So, how can I prevent dnsmasq from returning the parent domain address
when I look up <anyOldRubbish>.isbd.net ?  Presumably it can't find the
name locally and sends off the request to the upstream name server
which (sort of correctly) returns 195.74.61.93.

However it means that if I mis-type a name or if one of my machines
dies then I may not notice immediately because DNS still succeeds.

-- 
Chris Green



From justin.mcateer at gmail.com  Sun Nov  8 15:16:12 2009
From: justin.mcateer at gmail.com (Justin McAteer)
Date: Sun, 8 Nov 2009 09:16:12 -0600
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <cc73a6dc0911061740s1005632j5d745fa51b19d115@mail.gmail.com>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
	<cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>
	<cc73a6dc0911061740s1005632j5d745fa51b19d115@mail.gmail.com>
Message-ID: <39b4dffd0911080716m77615f75ne4dfb7b97a17f9e0@mail.gmail.com>

HS,

The phone will have to request option 120 from the server or DNSMasq
will not send the option. Alternatively, you may use dhcp-option-force
to have the server send the option whether the client requests it or
not.

Thank You,
Justin McAteer




On Fri, Nov 6, 2009 at 7:40 PM, H. S. <hs.samix at gmail.com> wrote:
> On Fri, Nov 6, 2009 at 8:30 PM, H. S. <hs.samix at gmail.com> wrote:
>> On Sun, Nov 1, 2009 at 4:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
>>
>>>
>>> Current version of dnsmasq know about the peculiar format of option 120, so
>>> you can just do:
>>>
>>> dhcp-option=120,192.168.5.1
>>
>> dnsmasq had no problem with this. Yet to try with the Nokia phone.
>
> Despite giving that option in dnsmasq and restarting it, the phone
> still did not grab a DHCP offer :(
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From perette at barella.org  Sun Nov  8 15:25:11 2009
From: perette at barella.org (Perette Barella)
Date: Sun, 8 Nov 2009 10:25:11 -0500
Subject: [Dnsmasq-discuss] dnsmasq returns addresses for non-existent
	hosts - what have I	mis-configured?
In-Reply-To: <20091108120803.GA14910@chris>
References: <20091108120803.GA14910@chris>
Message-ID: <EE23458F-184E-404B-AB8B-B9F7B6047FAB@barella.org>

It looks like your provider has set up a wildcard A record, which is  
similar to DNS hijacking as a "helpful" feature to users who miskey a  
domain name.  It's not isolated to you:

mugenshi:etc x10$ host ghijk.isbd.net
ghijk.isbd.net has address 195.74.61.93
ghijk.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.

You could check Gradwell's support pages, but I doubt there is an  
option to shut it off, since the DNS is published this way.  It's a  
publication problem/"feature", not a bug in dnsmasq.

Dnsmasq does have a a bogus-nxdomain option, which you give some IP  
addresses and those addresses are translated into NXDOMAIN non- 
existent domain responses.  However, if isbd.net (without leading  
characters) is actually a useful server to you, this solution is not  
viable because the valid use shares the same IP as the bogus responses.

Perette



On 2009?11?08?, at 7:08, Chris G wrote:

> I have dnsmasq working quite happily on a Ubuntu Server 9.10 system
> providing dns for my small SoHo network.
>
> I have just noticed however that if I ask for the address of a
> non-existent name dnsmasq returns the name of one of my hosting
> service's machines.  It always returns the same address for any
> non-existent name, e.g.:-
>
>    chris$ host abcde.isbd.net
>    abcde.isbd.net has address 195.74.61.93
>    abcde.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
>    chris$ host xyz
>    xyz.isbd.net has address 195.74.61.93
>    xyz.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
>    chris$ host xyz.isbd.net
>    xyz.isbd.net has address 195.74.61.93
>    xyz.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
>
>
> I have isbd.net and isbd.co.uk hosted at Gradwell so I do have a
> connection with them.  In fact things are becoming clearer now,
> 195.74.61.93 is the (quite correct) address returned when you
> look up isbd.net.
>
> So, how can I prevent dnsmasq from returning the parent domain address
> when I look up <anyOldRubbish>.isbd.net ?  Presumably it can't find  
> the
> name locally and sends off the request to the upstream name server
> which (sort of correctly) returns 195.74.61.93.
>
> However it means that if I mis-type a name or if one of my machines
> dies then I may not notice immediately because DNS still succeeds.
>
> -- 
> Chris Green
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss



From rob0 at gmx.co.uk  Sun Nov  8 15:37:36 2009
From: rob0 at gmx.co.uk (/dev/rob0)
Date: Sun, 8 Nov 2009 09:37:36 -0600
Subject: [Dnsmasq-discuss] dnsmasq returns addresses for non-existent
 hosts - what have I mis-configured?
In-Reply-To: <20091108120803.GA14910@chris>
References: <20091108120803.GA14910@chris>
Message-ID: <20091108153715.GB24728@minipax>

On Sun, Nov 08, 2009 at 12:08:04PM +0000, Chris G wrote:
> So, how can I prevent dnsmasq from returning the parent domain address
> when I look up <anyOldRubbish>.isbd.net ?  Presumably it can't find the

See "--local" in the man page.

> name locally and sends off the request to the upstream name server
> which (sort of correctly) returns 195.74.61.93.

NXDOMAIN hijacking is bad, a stupid idea by greedy people who want
more Web traffic, but who do not know how DNS is supposed to work.
See "--bogus-nxdomain" in the man page. Or better yet, find a better
upstream nameserver to use.

I run my own named(8) caching resolver on an alternate port for
dnsmasq to use. It's pretty simple, using --server and --no-resolv
settings. I changed the ports to non-privileged ones, so my named
can be started by and run as a non-root user.
-- 
    Offlist mail to this address is discarded unless
    "/dev/rob0" or "not-spam" is in Subject: header


From rob0 at gmx.co.uk  Sun Nov  8 22:53:02 2009
From: rob0 at gmx.co.uk (/dev/rob0)
Date: Sun, 8 Nov 2009 17:53:02 -0500
Subject: [Dnsmasq-discuss] server=... remote reverse DNS fails
Message-ID: <200911081653.02760.rob0@gmx.co.uk>

I have a local dnsmasq and another one connected via VPN. The local
names and DHCP'ed IP addresses resolve at each place. Each is set up
to refer to the other's forward and reverse zone. It seems to work
fully going one way, but only for the forward zone from the other.

Cast of Characters:
+-----------------+
192.168.8.97   Chestnut.pirate.lan, slamd64 [1] 12.2
192.168.4.9    gazoo.lan, openwrt Kamikaze 8.09.1 (brcm-2.4)
192.168.4.173  sip.gazoo.lan, Sipura SPA-2000
192.168.3.1    Chestnut's OpenVPN IP (the server)
192.168.3.8    gazoo's VPN IP 

Routing works. Chestnut is the openvpn server, and hosts in gazoo.lan
can get to pirate.lan hosts and to other openvpn clients elsewhere. I
noticed the problem when setting up gazoo and sip to do remote syslog
to chestnut; gazoo appears in logs as "gazoo", sip by IP.

Testing from a host in pirate.lan:

$ host 192.168.4.173
Host 173.4.168.192.in-addr.arpa. not found: 3(NXDOMAIN)
$ host 192.168.4.173 192.168.4.9
Using domain server:
Name: 192.168.4.9
Address: 192.168.4.9#53
Aliases:

173.4.168.192.in-addr.arpa domain name pointer sip.gazoo.lan.
$ host sip.gazoo.lan.
sip.lan has address 192.168.4.173

The dnsmasq.d directory on chestnut.pirate.lan has a file "gazoo",
with this:

server=/gazoo.lan/192.168.3.8
server=/4.168.192.in-addr.arpa/192.168.3.8

192.168.3.8 is the VPN IP address for the remote dnsmasq, which is
a/k/a 192.168.4.9 . The gazoo.lan names resolve, but reverse doesn't
work, and test results are the same with either of the VPN or LAN IP
addresses.

Continuing tests, we see that both forward and reverse work from the
other side:

$ host 192.168.8.172
172.8.168.192.in-addr.arpa domain name pointer Wii.pirate.lan.
$ host 192.168.8.172 192.168.4.9
Using domain server:
Name: 192.168.4.9
Address: 192.168.4.9#53
Aliases:

172.8.168.192.in-addr.arpa domain name pointer Wii.pirate.lan.
$ host Wii.pirate.lan.
Wii.pirate.lan has address 192.168.8.172
ws at whn:~$ host Wii.pirate.lan. 192.168.4.9
Using domain server:
Name: 192.168.4.9
Address: 192.168.4.9#53
Aliases:

Wii.pirate.lan has address 192.168.8.172

The dnsmasq which is not working properly, Slamd64 12.2:

root at chestnut:~# dnsmasq --version
Dnsmasq version 2.46  Copyright (C) 2000-2008 Simon Kelley
Compile time options IPv6 GNU-getopt no-DBus I18N TFTP

The one which is working, openwrt Kamikaze 8.09.1:

root at gazoo:~# dnsmasq --version
Dnsmasq version 2.47  Copyright (C) 2000-2009 Simon Kelley
Compile time options IPv6 GNU-getopt no-DBus no-I18N TFTP

I'll try upgrading and report back if that helps. Oh, here's another
test:

$ dig -x +trace 192.168.4.173

; <<>> DiG 9.4.2-P2 <<>> -x +trace 192.168.4.173
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 28788
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0

;; QUESTION SECTION:
;+trace.in-addr.arpa.           IN      PTR

;; AUTHORITY SECTION:
in-addr.arpa.           10800   IN      SOA     A.ROOT-SERVERS.NET. 
dns-ops.ARIN.NET. 2009110804 1800 900 691200 10800

;; Query time: 166 msec
;; SERVER: 192.168.8.97#53(192.168.8.97)
;; WHEN: Sun Nov  8 15:45:37 2009
;; MSG SIZE  rcvd: 104

;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 31979
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;192.168.4.173.                 IN      A

;; ANSWER SECTION:
192.168.4.173.          7200    IN      A       192.168.4.173

;; Query time: 2 msec
;; SERVER: 192.168.8.97#53(192.168.8.97)
;; WHEN: Sun Nov  8 15:45:37 2009
;; MSG SIZE  rcvd: 47

Note, all of these tests were done from clients on the pirate.lan
side. ATM there's no dig(1) on the gazoo.lan side, and gazoo itself
lacks storage capacity to install it. (The busybox implementation of
nslookup(1) is even worse than BIND's.) I think the directed queries

Curious that there would be an A query and answer. Using ptr for the
in-addr.arpa. name rather than -x, it appears to be ignoring the
server declaration for the reverse zone:

$ dig +trace 173.4.168.192.in-addr.arpa. ptr

; <<>> DiG 9.4.2-P2 <<>> +trace 173.4.168.192.in-addr.arpa. ptr
;; global options:  printcmd
.                       372212  IN      NS      G.ROOT-SERVERS.NET.
.                       372212  IN      NS      B.ROOT-SERVERS.NET.
.                       372212  IN      NS      A.ROOT-SERVERS.NET.
.                       372212  IN      NS      F.ROOT-SERVERS.NET.
.                       372212  IN      NS      I.ROOT-SERVERS.NET.
.                       372212  IN      NS      C.ROOT-SERVERS.NET.
.                       372212  IN      NS      H.ROOT-SERVERS.NET.
.                       372212  IN      NS      M.ROOT-SERVERS.NET.
.                       372212  IN      NS      J.ROOT-SERVERS.NET.
.                       372212  IN      NS      L.ROOT-SERVERS.NET.
.                       372212  IN      NS      D.ROOT-SERVERS.NET.
.                       372212  IN      NS      K.ROOT-SERVERS.NET.
.                       372212  IN      NS      E.ROOT-SERVERS.NET.
;; Received 512 bytes from 127.0.0.1#53(127.0.0.1) in 0 ms$ dig +trace 
173.4.168.192.in-addr.arpa. ptr

; <<>> DiG 9.4.2-P2 <<>> +trace 173.4.168.192.in-addr.arpa. ptr
;; global options:  printcmd
.                       372212  IN      NS      G.ROOT-SERVERS.NET.
.                       372212  IN      NS      B.ROOT-SERVERS.NET.
.                       372212  IN      NS      A.ROOT-SERVERS.NET.
.                       372212  IN      NS      F.ROOT-SERVERS.NET.
.                       372212  IN      NS      I.ROOT-SERVERS.NET.
.                       372212  IN      NS      C.ROOT-SERVERS.NET.
.                       372212  IN      NS      H.ROOT-SERVERS.NET.
.                       372212  IN      NS      M.ROOT-SERVERS.NET.
.                       372212  IN      NS      J.ROOT-SERVERS.NET.
.                       372212  IN      NS      L.ROOT-SERVERS.NET.
.                       372212  IN      NS      D.ROOT-SERVERS.NET.
.                       372212  IN      NS      K.ROOT-SERVERS.NET.
.                       372212  IN      NS      E.ROOT-SERVERS.NET.
;; Received 512 bytes from 127.0.0.1#53(127.0.0.1) in 0 ms

arpa.                   172800  IN      NS      D.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      F.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      G.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      B.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      E.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      M.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      A.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      I.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      H.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      C.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      L.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      K.ROOT-SERVERS.NET.
;; Received 496 bytes from 192.58.128.30#53(J.ROOT-SERVERS.NET) in 221 ms

192.in-addr.arpa.       86400   IN      NS      BASIL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      HENNA.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Z.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      X.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Y.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      DILL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      INDIGO.ARIN.NET.
;; Received 180 bytes from 128.8.10.90#53(D.ROOT-SERVERS.NET) in 96 ms

168.192.in-addr.arpa.   86400   IN      NS      blackhole-1.iana.org.
168.192.in-addr.arpa.   86400   IN      NS      blackhole-2.iana.org.
;; Received 104 bytes from 192.55.83.32#53(BASIL.ARIN.NET) in 208 ms

168.192.in-addr.arpa.   300     IN      SOA     prisoner.iana.org. 
hostmaster.root-servers.org. 2002040800 1800 900 604800 604800
;; Received 121 bytes from 192.175.48.6#53(blackhole-1.iana.org) in 106 ms

arpa.                   172800  IN      NS      D.ROOT-SERVERS.NET.$ dig +trace 
173.4.168.192.in-addr.arpa. ptr

; <<>> DiG 9.4.2-P2 <<>> +trace 173.4.168.192.in-addr.arpa. ptr
;; global options:  printcmd
.                       372212  IN      NS      G.ROOT-SERVERS.NET.
.                       372212  IN      NS      B.ROOT-SERVERS.NET.
.                       372212  IN      NS      A.ROOT-SERVERS.NET.
.                       372212  IN      NS      F.ROOT-SERVERS.NET.
.                       372212  IN      NS      I.ROOT-SERVERS.NET.
.                       372212  IN      NS      C.ROOT-SERVERS.NET.
.                       372212  IN      NS      H.ROOT-SERVERS.NET.
.                       372212  IN      NS      M.ROOT-SERVERS.NET.
.                       372212  IN      NS      J.ROOT-SERVERS.NET.
.                       372212  IN      NS      L.ROOT-SERVERS.NET.
.                       372212  IN      NS      D.ROOT-SERVERS.NET.
.                       372212  IN      NS      K.ROOT-SERVERS.NET.
.                       372212  IN      NS      E.ROOT-SERVERS.NET.
;; Received 512 bytes from 127.0.0.1#53(127.0.0.1) in 0 ms

arpa.                   172800  IN      NS      D.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      F.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      G.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      B.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      E.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      M.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      A.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      I.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      H.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      C.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      L.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      K.ROOT-SERVERS.NET.
;; Received 496 bytes from 192.58.128.30#53(J.ROOT-SERVERS.NET) in 221 ms

192.in-addr.arpa.       86400   IN      NS      BASIL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      HENNA.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Z.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      X.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Y.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      DILL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      INDIGO.ARIN.NET.
;; Received 180 bytes from 128.8.10.90#53(D.ROOT-SERVERS.NET) in 96 ms

168.192.in-addr.arpa.   86400   IN      NS      blackhole-1.iana.org.
168.192.in-addr.arpa.   86400   IN      NS      blackhole-2.iana.org.
;; Received 104 bytes from 192.55.83.32#53(BASIL.ARIN.NET) in 208 ms

168.192.in-addr.arpa.   300     IN      SOA     prisoner.iana.org. 
hostmaster.root-servers.org. 2002040800 1800 900 604800 604800
;; Received 121 bytes from 192.175.48.6#53(blackhole-1.iana.org) in 106 ms
arpa.                   172800  IN      NS      F.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      G.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      B.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      E.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      M.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      A.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      I.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      H.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      C.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      L.ROOT-SERVERS.NET.
arpa.                   172800  IN      NS      K.ROOT-SERVERS.NET.
;; Received 496 bytes from 192.58.128.30#53(J.ROOT-SERVERS.NET) in 221 ms

192.in-addr.arpa.       86400   IN      NS      BASIL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      HENNA.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Z.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      X.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      Y.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      DILL.ARIN.NET.
192.in-addr.arpa.       86400   IN      NS      INDIGO.ARIN.NET.
;; Received 180 bytes from 128.8.10.90#53(D.ROOT-SERVERS.NET) in 96 ms

168.192.in-addr.arpa.   86400   IN      NS      blackhole-1.iana.org.
168.192.in-addr.arpa.   86400   IN      NS      blackhole-2.iana.org.
;; Received 104 bytes from 192.55.83.32#53(BASIL.ARIN.NET) in 208 ms

168.192.in-addr.arpa.   300     IN      SOA     prisoner.iana.org. 
hostmaster.root-servers.org. 2002040800 1800 900 604800 604800
;; Received 121 bytes from 192.175.48.6#53(blackhole-1.iana.org) in 106 ms

This is all nothing but a minor annoyance for me, but I thought it
might be worth looking into.


[1] Slamd64 is/was a port of 32-bit Slackware to x86_64, probably in
    the process of being phased out since Slackware 13 had a x86_64
    version.
-- 
    Offlist mail to this address is discarded unless
    "/dev/rob0" or "not-spam" is in Subject: header


From cl at isbd.net  Mon Nov  9 13:58:25 2009
From: cl at isbd.net (Chris G)
Date: Mon, 9 Nov 2009 13:58:25 +0000
Subject: [Dnsmasq-discuss] dnsmasq returns addresses for non-existent
	hosts - what have I	mis-configured?
In-Reply-To: <EE23458F-184E-404B-AB8B-B9F7B6047FAB@barella.org>
References: <20091108120803.GA14910@chris>
	<EE23458F-184E-404B-AB8B-B9F7B6047FAB@barella.org>
Message-ID: <20091109135825.GC14910@chris>

On Sun, Nov 08, 2009 at 10:25:11AM -0500, Perette Barella wrote:
> It looks like your provider has set up a wildcard A record, which is  
> similar to DNS hijacking as a "helpful" feature to users who miskey a  
> domain name.  It's not isolated to you:
> 
> mugenshi:etc x10$ host ghijk.isbd.net
> ghijk.isbd.net has address 195.74.61.93
> ghijk.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
> 
> You could check Gradwell's support pages, but I doubt there is an  
> option to shut it off, since the DNS is published this way.  It's a  
> publication problem/"feature", not a bug in dnsmasq.
> 
I do in fact have the ability to change my domain's zone files.

... and there is what you describe (N.B. this from a web form, not
exact zone file syntax) :-

    *   195.74.61.93                    86400   A
    *   10 mail-in-1.lb.gradwell.net.   86400   MX

So can I simply delete these two entries?  (OK, people mis-typing domain
names *might* be affected but that's mostly me so I don't see a big
issue there)

-- 
Chris Green



From simon at thekelleys.org.uk  Mon Nov  9 21:27:58 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 09 Nov 2009 21:27:58 +0000
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>
References: <hcfnfk$326$1@ger.gmane.org> <4AEE0098.5060909@thekelleys.org.uk>
	<cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>
Message-ID: <4AF8895E.4020106@thekelleys.org.uk>


>> dhcp-option=option:sip,192.168.5.1
> 
> dnsmasq barfed at this:
>> sudo /etc/init.d/dnsmasq restart
> Restarting DNS forwarder and DHCP server: dnsmasq
> dnsmasq: bad dhcp-option at line 99 of /etc/dnsmasq.d/01-home-hosts.conf
>  failed!

Apologies, the option name is "sip-server", not "sip"

Simon.


From simon at thekelleys.org.uk  Mon Nov  9 21:30:18 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 09 Nov 2009 21:30:18 +0000
Subject: [Dnsmasq-discuss] dhcp-option 120, sip server, how?
In-Reply-To: <39b4dffd0911080716m77615f75ne4dfb7b97a17f9e0@mail.gmail.com>
References: <hcfnfk$326$1@ger.gmane.org>
	<4AEE0098.5060909@thekelleys.org.uk>	<cc73a6dc0911061730g1cc515dco574728d1b1286f57@mail.gmail.com>	<cc73a6dc0911061740s1005632j5d745fa51b19d115@mail.gmail.com>
	<39b4dffd0911080716m77615f75ne4dfb7b97a17f9e0@mail.gmail.com>
Message-ID: <4AF889EA.9040703@thekelleys.org.uk>

Setting log-dhcp in /etc/dnsmasq.conf will give you useful information 
about exactly what is going on here.

Simon.

Justin McAteer wrote:
> HS,
> 
> The phone will have to request option 120 from the server or DNSMasq
> will not send the option. Alternatively, you may use dhcp-option-force
> to have the server send the option whether the client requests it or
> not.
> 
> Thank You,
> Justin McAteer
> 
> 
> 
> 
> On Fri, Nov 6, 2009 at 7:40 PM, H. S. <hs.samix at gmail.com> wrote:
>> On Fri, Nov 6, 2009 at 8:30 PM, H. S. <hs.samix at gmail.com> wrote:
>>> On Sun, Nov 1, 2009 at 4:41 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
>>>
>>>> Current version of dnsmasq know about the peculiar format of option 120, so
>>>> you can just do:
>>>>
>>>> dhcp-option=120,192.168.5.1
>>> dnsmasq had no problem with this. Yet to try with the Nokia phone.
>> Despite giving that option in dnsmasq and restarting it, the phone
>> still did not grab a DHCP offer :(
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 



From simon at thekelleys.org.uk  Mon Nov  9 21:39:30 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 09 Nov 2009 21:39:30 +0000
Subject: [Dnsmasq-discuss] server=... remote reverse DNS fails
In-Reply-To: <200911081653.02760.rob0@gmx.co.uk>
References: <200911081653.02760.rob0@gmx.co.uk>
Message-ID: <4AF88C12.30103@thekelleys.org.uk>

/dev/rob0 wrote:

Snip lots of tests.

> This is all nothing but a minor annoyance for me, but I thought it
> might be worth looking into.
> 
> 

I'm not aware of anything that could cause this. The best way to see 
what's going on might be to set --log-queries on both dnsmasq instances.


Cheers,

Simon


From gerrit at pmp.uni-hannover.de  Wed Nov 11 10:13:01 2009
From: gerrit at pmp.uni-hannover.de (Gerrit =?ISO-8859-1?Q?K=FChn?=)
Date: Wed, 11 Nov 2009 11:13:01 +0100
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts
 in its own domain
In-Reply-To: <4AF1684F.8060203@thekelleys.org.uk>
References: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>
	<4AF1684F.8060203@thekelleys.org.uk>
Message-ID: <20091111111301.b85cfe7b.gerrit@pmp.uni-hannover.de>

On Wed, 04 Nov 2009 11:41:03 +0000 Simon Kelley <simon at thekelleys.org.uk>
wrote about Re: [Dnsmasq-discuss] dnsmasq forwards requests for unknown
hosts in its own domain:

SK> > Does anyone here have a hint for me how to setup this correctly and
SK> > how to prevent queries for unknown hosts being sent back from the
SK> > intranet dnsmasq?

SK> local=/intranet.my.domain/
SK> on the intranet server will stop it forwarding any requests for that 
SK> domain. I think that's what you need.

This fixed the directly triggerable ping-pong between the two servers. Now
I still see these requests sometimes:

Nov 11 10:30:19 theq dnsmasq[97791]: query[PTR]
db._dns-sd._udp.0.64.168.192.in- addr.arpa from 10.117.15.254
Nov 11 10:30:19 theq dnsmasq[97791]: forwarded
db._dns-sd._udp.0.64.168.192.in-a ddr.arpa to 10.117.15.254

These requests/forwards won't stop and put quite some load on dnsmasq
running on 10.117.15.254. After stopping, waiting some seconds and
starting again dnsmasq on 192.168.64.254, everything is back to normal
(for a day or so, until this starts again).

Can anyone explain to me what might cause these requests and how I can
prevent them either from being requested at all or at least from being
forwarded?


cu
  Gerrit


From gerrit at pmp.uni-hannover.de  Wed Nov 11 10:28:58 2009
From: gerrit at pmp.uni-hannover.de (Gerrit =?ISO-8859-1?Q?K=FChn?=)
Date: Wed, 11 Nov 2009 11:28:58 +0100
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts
 in its own domain
In-Reply-To: <20091111111301.b85cfe7b.gerrit@pmp.uni-hannover.de>
References: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>
	<4AF1684F.8060203@thekelleys.org.uk>
	<20091111111301.b85cfe7b.gerrit@pmp.uni-hannover.de>
Message-ID: <20091111112858.065e85c4.gerrit@pmp.uni-hannover.de>

On Wed, 11 Nov 2009 11:13:01 +0100 Gerrit K?hn
<gerrit at pmp.uni-hannover.de> wrote about Re: [Dnsmasq-discuss] dnsmasq
forwards requests for unknown hosts in its own domain:

GK> Nov 11 10:30:19 theq dnsmasq[97791]: query[PTR]
GK> db._dns-sd._udp.0.64.168.192.in- addr.arpa from 10.117.15.254
GK> Nov 11 10:30:19 theq dnsmasq[97791]: forwarded
GK> db._dns-sd._udp.0.64.168.192.in-a ddr.arpa to 10.117.15.254

GK> These requests/forwards won't stop and put quite some load on dnsmasq
GK> running on 10.117.15.254. 

Sorry, wrong hostname here. To make the situation clear:
dnsmasq is running on two hosts:

theq (10.117.1.254) <-> nat-router (10.117.15.254 / 192.168.64.254)

I do not know where the requests for 192.168.64.0 stem from, but dnsmasq
running on the nat-router forwards them to dnsmasq on theq.


cu
  Gerrit


From 4lists at gmail.com  Wed Nov 11 11:48:48 2009
From: 4lists at gmail.com (Edson - Lists)
Date: Wed, 11 Nov 2009 09:48:48 -0200
Subject: [Dnsmasq-discuss] dnsmasq forwards requests for unknown hosts
 in its own domain
In-Reply-To: <20091111112858.065e85c4.gerrit@pmp.uni-hannover.de>
References: <20091104123503.179fc6ca.gerrit@pmp.uni-hannover.de>	<4AF1684F.8060203@thekelleys.org.uk>	<20091111111301.b85cfe7b.gerrit@pmp.uni-hannover.de>
	<20091111112858.065e85c4.gerrit@pmp.uni-hannover.de>
Message-ID: <4AFAA4A0.1000203@gmail.com>

Maybe this hints could help You:
1) make sure that dnsmasq on nat-router is listening on the right 
interfaces (just internal?)
2) arm syslog for all DNS queries and see from where they are coming.

Edson.

Gerrit K?hn escreveu:
> On Wed, 11 Nov 2009 11:13:01 +0100 Gerrit K?hn
> <gerrit at pmp.uni-hannover.de> wrote about Re: [Dnsmasq-discuss] dnsmasq
> forwards requests for unknown hosts in its own domain:
> 
> GK> Nov 11 10:30:19 theq dnsmasq[97791]: query[PTR]
> GK> db._dns-sd._udp.0.64.168.192.in- addr.arpa from 10.117.15.254
> GK> Nov 11 10:30:19 theq dnsmasq[97791]: forwarded
> GK> db._dns-sd._udp.0.64.168.192.in-a ddr.arpa to 10.117.15.254
> 
> GK> These requests/forwards won't stop and put quite some load on dnsmasq
> GK> running on 10.117.15.254. 
> 
> Sorry, wrong hostname here. To make the situation clear:
> dnsmasq is running on two hosts:
> 
> theq (10.117.1.254) <-> nat-router (10.117.15.254 / 192.168.64.254)
> 
> I do not know where the requests for 192.168.64.0 stem from, but dnsmasq
> running on the nat-router forwards them to dnsmasq on theq.
> 
> 
> cu
>   Gerrit
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 


From knud.skrald at guldberg.info  Sat Nov 14 23:34:19 2009
From: knud.skrald at guldberg.info (knud.skrald at guldberg.info)
Date: Sun, 15 Nov 2009 00:34:19 +0100
Subject: [Dnsmasq-discuss] can resolve 1 specific host
Message-ID: <4AFF3E7B.302@guldberg.info>

Hi

 From time to time I get problmem with one host name
www.dba.dk and camping.dba.dk

They can not be solved, seems to be locked in some way on my dnsmasq server.

A snip (reduced) of my log file is attached.

Knud
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: log1.txt
Url: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091115/f28a18c5/attachment.txt 

From cl at isbd.net  Sun Nov 15 10:44:12 2009
From: cl at isbd.net (Chris G)
Date: Sun, 15 Nov 2009 10:44:12 +0000
Subject: [Dnsmasq-discuss] dnsmasq returns addresses for non-existent
	hosts - what have I	mis-configured?
In-Reply-To: <20091109135825.GC14910@chris>
References: <20091108120803.GA14910@chris>
	<EE23458F-184E-404B-AB8B-B9F7B6047FAB@barella.org>
	<20091109135825.GC14910@chris>
Message-ID: <20091115104412.GG15230@chris>

On Mon, Nov 09, 2009 at 01:58:25PM +0000, Chris G wrote:
> On Sun, Nov 08, 2009 at 10:25:11AM -0500, Perette Barella wrote:
> > It looks like your provider has set up a wildcard A record, which is  
> > similar to DNS hijacking as a "helpful" feature to users who miskey a  
> > domain name.  It's not isolated to you:
> > 
> > mugenshi:etc x10$ host ghijk.isbd.net
> > ghijk.isbd.net has address 195.74.61.93
> > ghijk.isbd.net mail is handled by 10 mail-in-1.lb.gradwell.net.
> > 
> > You could check Gradwell's support pages, but I doubt there is an  
> > option to shut it off, since the DNS is published this way.  It's a  
> > publication problem/"feature", not a bug in dnsmasq.
> > 
> I do in fact have the ability to change my domain's zone files.
> 
> ... and there is what you describe (N.B. this from a web form, not
> exact zone file syntax) :-
> 
>     *   195.74.61.93                    86400   A
>     *   10 mail-in-1.lb.gradwell.net.   86400   MX
> 
> So can I simply delete these two entries?  (OK, people mis-typing domain
> names *might* be affected but that's mostly me so I don't see a big
> issue there)
> 
... and the answer is yes.  I've deleted the wild card lines from the
form and now <non-existent>.isbd.net returns "not found", excellent!

Thanks for all the help here.

-- 
Chris Green



From simon at thekelleys.org.uk  Sun Nov 15 12:18:58 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Sun, 15 Nov 2009 12:18:58 +0000
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4AFF3E7B.302@guldberg.info>
References: <4AFF3E7B.302@guldberg.info>
Message-ID: <4AFFF1B2.8050505@thekelleys.org.uk>

knud.skrald at guldberg.info wrote:
> Hi
> 
> From time to time I get problmem with one host name
> www.dba.dk and camping.dba.dk
> 
> They can not be solved, seems to be locked in some way on my dnsmasq
> server.
> 
> A snip (reduced) of my log file is attached.
> 
> Knud
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss

Nov 14 20:58:46 dnsmasq[21682]: query[A] www.dba.dk from 10.2.1.182
Nov 14 20:58:46 dnsmasq[21682]: forwarded www.dba.dk to 89.184.128.193
Nov 14 20:58:46 dnsmasq[21682]: reply www.dba.dk is 91.213.17.20

That looks good.....


Nov 14 20:58:59 dnsmasq[21682]: query[A] camping.dba.dk from 10.2.1.182
Nov 14 20:58:59 dnsmasq[21682]: forwarded camping.dba.dk to 89.184.128.193
Nov 14 20:58:59 dnsmasq[21682]: reply camping.dba.dk is <CNAME>
Nov 14 20:58:59 dnsmasq[21682]: reply www.dba.dk is NODATA-IPv4

Now we look up camping.dba.dk, and get back the answer that
camping.dba.dk is a CNAME for www.dba.dk, and no data for www.dba.dk.
This means that the domain www.dba.dk exists, but there is no "A" record
for it. That information is cached, so future attemps to look up
www.dba.dk or camping.dba.dk fail, because dnsmasq "knows" that
www.dba.dk has no data.

There's clearly something wrong in the upstream server at
89.184.128.193, since it gave the correct data to the first request.

I've tried a query to 89.184.128.193 and I get the correct information,
so it could be a problem which has been resolved:

srk at spike$ dig @89.184.128.193 camping.dba.dk

; <<>> DiG 9.5.1-P3 <<>> @89.184.128.193 camping.dba.dk
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 19814
;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 4, ADDITIONAL: 4

;; QUESTION SECTION:
;camping.dba.dk.			IN	A

;; ANSWER SECTION:
camping.dba.dk.		600	IN	CNAME	www.dba.dk.
www.dba.dk.		1444	IN	A	91.213.17.20

running this query through my local dnsmasq causes the correct thing to
be cached.


Summary: the problem seems to be in the upstream nameserver, but it's OK
now. If you are still hitting problems, try restarting dnsmasq to clear
out the bad cache entries, and try again: If it's still broken, complain
to whoever runs that nameserver.

Cheers,

Simon.



From knud.skrald at guldberg.info  Sun Nov 15 13:29:11 2009
From: knud.skrald at guldberg.info (Knud)
Date: Sun, 15 Nov 2009 14:29:11 +0100
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4AFFF1B2.8050505@thekelleys.org.uk>
References: <4AFF3E7B.302@guldberg.info> <4AFFF1B2.8050505@thekelleys.org.uk>
Message-ID: <4B000227.9010205@guldberg.info>

Hi Simon

Just tested again...same issue again

asking my upstream nameserver...
camping.dba.dk => unknown
www.dba.dk => ok
again camping.dba.dk => now ok !

restart my dnsmasq makes it to known www.dba.dk and camping.dba.dk

I am running another dnsmasq server (at a school) with another upstream 
nameserver...there seems no problems
It is another version of dnsmasq included in the FREESCO packged.
But it seems not to be a dnsmasq issue but further upstream problem

Just strange that this specific host gives problems and not others (to 
my knowlegde)

Knud



Simon Kelley wrote:
> knud.skrald at guldberg.info wrote:
>> Hi
>>
>> From time to time I get problmem with one host name
>> www.dba.dk and camping.dba.dk
>>
>> They can not be solved, seems to be locked in some way on my dnsmasq
>> server.
>>
>> A snip (reduced) of my log file is attached.
>>
>> Knud
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 
> Nov 14 20:58:46 dnsmasq[21682]: query[A] www.dba.dk from 10.2.1.182
> Nov 14 20:58:46 dnsmasq[21682]: forwarded www.dba.dk to 89.184.128.193
> Nov 14 20:58:46 dnsmasq[21682]: reply www.dba.dk is 91.213.17.20
> 
> That looks good.....
> 
> 
> Nov 14 20:58:59 dnsmasq[21682]: query[A] camping.dba.dk from 10.2.1.182
> Nov 14 20:58:59 dnsmasq[21682]: forwarded camping.dba.dk to 89.184.128.193
> Nov 14 20:58:59 dnsmasq[21682]: reply camping.dba.dk is <CNAME>
> Nov 14 20:58:59 dnsmasq[21682]: reply www.dba.dk is NODATA-IPv4
> 
> Now we look up camping.dba.dk, and get back the answer that
> camping.dba.dk is a CNAME for www.dba.dk, and no data for www.dba.dk.
> This means that the domain www.dba.dk exists, but there is no "A" record
> for it. That information is cached, so future attemps to look up
> www.dba.dk or camping.dba.dk fail, because dnsmasq "knows" that
> www.dba.dk has no data.
> 
> There's clearly something wrong in the upstream server at
> 89.184.128.193, since it gave the correct data to the first request.
> 
> I've tried a query to 89.184.128.193 and I get the correct information,
> so it could be a problem which has been resolved:
> 
> srk at spike$ dig @89.184.128.193 camping.dba.dk
> 
> ; <<>> DiG 9.5.1-P3 <<>> @89.184.128.193 camping.dba.dk
> ; (1 server found)
> ;; global options:  printcmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 19814
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 4, ADDITIONAL: 4
> 
> ;; QUESTION SECTION:
> ;camping.dba.dk.			IN	A
> 
> ;; ANSWER SECTION:
> camping.dba.dk.		600	IN	CNAME	www.dba.dk.
> www.dba.dk.		1444	IN	A	91.213.17.20
> 
> running this query through my local dnsmasq causes the correct thing to
> be cached.
> 
> 
> Summary: the problem seems to be in the upstream nameserver, but it's OK
> now. If you are still hitting problems, try restarting dnsmasq to clear
> out the bad cache entries, and try again: If it's still broken, complain
> to whoever runs that nameserver.
> 
> Cheers,
> 
> Simon.
> 
> 



From simon at thekelleys.org.uk  Sun Nov 15 15:32:57 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Sun, 15 Nov 2009 15:32:57 +0000
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4B000227.9010205@guldberg.info>
References: <4AFF3E7B.302@guldberg.info> <4AFFF1B2.8050505@thekelleys.org.uk>
	<4B000227.9010205@guldberg.info>
Message-ID: <4B001F29.8010806@thekelleys.org.uk>

Knud wrote:
> Hi Simon
> 
> Just tested again...same issue again
> 
> asking my upstream nameserver...
> camping.dba.dk => unknown
> www.dba.dk => ok
> again camping.dba.dk => now ok !

Yes, I tried again and saw the same thing.

;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;camping.dba.dk.			IN	A

;; ANSWER SECTION:
camping.dba.dk.		600	IN	CNAME	www.dba.dk.

;; Query time: 77 msec


repeat:

;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 4, ADDITIONAL: 4

;; QUESTION SECTION:
;camping.dba.dk.			IN	A

;; ANSWER SECTION:
camping.dba.dk.		585	IN	CNAME	www.dba.dk.
www.dba.dk.		1787	IN	A	91.213.17.20

;; AUTHORITY SECTION:
dba.dk.			21587	IN	NS	ns6.netnames.net.
dba.dk.			21587	IN	NS	ns1.netnames.net.
dba.dk.			21587	IN	NS	ns5.netnames.net.
dba.dk.			21587	IN	NS	ns2.netnames.net.

;; ADDITIONAL SECTION:
ns6.netnames.net.	168883	IN	A	199.7.69.253
ns1.netnames.net.	21747	IN	A	204.74.108.253
ns5.netnames.net.	168883	IN	A	199.7.68.253
ns2.netnames.net.	168883	IN	A	62.128.146.40

;; Query time: 46 msec


Looks like once the upstream server has the CNAME cached, it works OK,
but not before. There are more clues there, as it tells us the
authoritative nameservers (Ie the ones that hold the configured data)
are a netnames.net.

We can ask one of those direct:

srk at spike$ dig @ns6.netnames.net camping.dba.dk

; <<>> DiG 9.5.1-P3 <<>> @ns6.netnames.net camping.dba.dk
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 10510
;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;camping.dba.dk.			IN	A

;; ANSWER SECTION:
camping.dba.dk.		600	IN	CNAME	www.dba.dk.

;; Query time: 22 msec
;; SERVER: 199.7.69.253#53(199.7.69.253)
;; WHEN: Sun Nov 15 15:20:05 2009
;; MSG SIZE  rcvd: 50


Which probably shows a real problem. That nameserver always gives that
data without the A record.


But even more strange, the netnames.net nameservers are not the real
authoritative ones for dba.dk

srk at spike:/mnt/onetouch/simon/torrents$ dig ns dba.dk

; <<>> DiG 9.5.1-P3 <<>> ns dba.dk
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 46594
;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 2

;; QUESTION SECTION:
;dba.dk.				IN	NS

;; ANSWER SECTION:
dba.dk.			43888	IN	NS	ns1.danhost.dk.
dba.dk.			43888	IN	NS	ns0.danhost.dk.

;; ADDITIONAL SECTION:
ns0.danhost.dk.		14276	IN	A	217.195.182.61
ns1.danhost.dk.		13926	IN	A	195.140.132.26

;; Query time: 9 msec
;; SERVER: 192.168.0.4#53(192.168.0.4)
;; WHEN: Sun Nov 15 15:27:37 2009
;; MSG SIZE  rcvd: 100

so the 89.184.128.193 nameserver you are using is very confused, I
think. It's using the wrong authoritative nameservers for dba.dk and
getting the wrong answers, which it is passing on. At this point it's
probably best to pass the whole mess over to the admins of
89.184.128.193 (your ISP?) CC: me on the mail if you wish, in case I can
provide more information.

> 
> restart my dnsmasq makes it to known www.dba.dk and camping.dba.dk
> 
> I am running another dnsmasq server (at a school) with another upstream
> nameserver...there seems no problems

I see the same: there's something wrong with 89.184.128.193, for sure.

> It is another version of dnsmasq included in the FREESCO packged.
> But it seems not to be a dnsmasq issue but further upstream problem
> 
> Just strange that this specific host gives problems and not others (to
> my knowlegde)


Cheers,

Simon.


From knud.skrald at guldberg.info  Sun Nov 15 16:06:02 2009
From: knud.skrald at guldberg.info (Knud)
Date: Sun, 15 Nov 2009 17:06:02 +0100
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4B001F29.8010806@thekelleys.org.uk>
References: <4AFF3E7B.302@guldberg.info> <4AFFF1B2.8050505@thekelleys.org.uk>
	<4B000227.9010205@guldberg.info>
	<4B001F29.8010806@thekelleys.org.uk>
Message-ID: <4B0026EA.60002@guldberg.info>

Hi Simon

I have opened a support ticket at my ISP (you right..)
Informed them about the strange behavoir of there NS
Let us see what happends

Otherwise they have been very stable...it's a fiberoptic connection 
25/25 Mbit.

Knud

Simon Kelley wrote:
> Knud wrote:
>> Hi Simon
>>
>> Just tested again...same issue again
>>
>> asking my upstream nameserver...
>> camping.dba.dk => unknown
>> www.dba.dk => ok
>> again camping.dba.dk => now ok !
> 
> Yes, I tried again and saw the same thing.
> 
> ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
> 
> ;; QUESTION SECTION:
> ;camping.dba.dk.			IN	A
> 
> ;; ANSWER SECTION:
> camping.dba.dk.		600	IN	CNAME	www.dba.dk.
> 
> ;; Query time: 77 msec
> 
> 
> repeat:
> 
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 4, ADDITIONAL: 4
> 
> ;; QUESTION SECTION:
> ;camping.dba.dk.			IN	A
> 
> ;; ANSWER SECTION:
> camping.dba.dk.		585	IN	CNAME	www.dba.dk.
> www.dba.dk.		1787	IN	A	91.213.17.20
> 
> ;; AUTHORITY SECTION:
> dba.dk.			21587	IN	NS	ns6.netnames.net.
> dba.dk.			21587	IN	NS	ns1.netnames.net.
> dba.dk.			21587	IN	NS	ns5.netnames.net.
> dba.dk.			21587	IN	NS	ns2.netnames.net.
> 
> ;; ADDITIONAL SECTION:
> ns6.netnames.net.	168883	IN	A	199.7.69.253
> ns1.netnames.net.	21747	IN	A	204.74.108.253
> ns5.netnames.net.	168883	IN	A	199.7.68.253
> ns2.netnames.net.	168883	IN	A	62.128.146.40
> 
> ;; Query time: 46 msec
> 
> 
> Looks like once the upstream server has the CNAME cached, it works OK,
> but not before. There are more clues there, as it tells us the
> authoritative nameservers (Ie the ones that hold the configured data)
> are a netnames.net.
> 
> We can ask one of those direct:
> 
> srk at spike$ dig @ns6.netnames.net camping.dba.dk
> 
> ; <<>> DiG 9.5.1-P3 <<>> @ns6.netnames.net camping.dba.dk
> ; (1 server found)
> ;; global options:  printcmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 10510
> ;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
> ;; WARNING: recursion requested but not available
> 
> ;; QUESTION SECTION:
> ;camping.dba.dk.			IN	A
> 
> ;; ANSWER SECTION:
> camping.dba.dk.		600	IN	CNAME	www.dba.dk.
> 
> ;; Query time: 22 msec
> ;; SERVER: 199.7.69.253#53(199.7.69.253)
> ;; WHEN: Sun Nov 15 15:20:05 2009
> ;; MSG SIZE  rcvd: 50
> 
> 
> Which probably shows a real problem. That nameserver always gives that
> data without the A record.
> 
> 
> But even more strange, the netnames.net nameservers are not the real
> authoritative ones for dba.dk
> 
> srk at spike:/mnt/onetouch/simon/torrents$ dig ns dba.dk
> 
> ; <<>> DiG 9.5.1-P3 <<>> ns dba.dk
> ;; global options:  printcmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 46594
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 2
> 
> ;; QUESTION SECTION:
> ;dba.dk.				IN	NS
> 
> ;; ANSWER SECTION:
> dba.dk.			43888	IN	NS	ns1.danhost.dk.
> dba.dk.			43888	IN	NS	ns0.danhost.dk.
> 
> ;; ADDITIONAL SECTION:
> ns0.danhost.dk.		14276	IN	A	217.195.182.61
> ns1.danhost.dk.		13926	IN	A	195.140.132.26
> 
> ;; Query time: 9 msec
> ;; SERVER: 192.168.0.4#53(192.168.0.4)
> ;; WHEN: Sun Nov 15 15:27:37 2009
> ;; MSG SIZE  rcvd: 100
> 
> so the 89.184.128.193 nameserver you are using is very confused, I
> think. It's using the wrong authoritative nameservers for dba.dk and
> getting the wrong answers, which it is passing on. At this point it's
> probably best to pass the whole mess over to the admins of
> 89.184.128.193 (your ISP?) CC: me on the mail if you wish, in case I can
> provide more information.
> 
>> restart my dnsmasq makes it to known www.dba.dk and camping.dba.dk
>>
>> I am running another dnsmasq server (at a school) with another upstream
>> nameserver...there seems no problems
> 
> I see the same: there's something wrong with 89.184.128.193, for sure.
> 
>> It is another version of dnsmasq included in the FREESCO packged.
>> But it seems not to be a dnsmasq issue but further upstream problem
>>
>> Just strange that this specific host gives problems and not others (to
>> my knowlegde)
> 
> 
> Cheers,
> 
> Simon.
> 



From cscott at agencypja.com  Mon Nov 16 20:10:00 2009
From: cscott at agencypja.com (Christopher Scott)
Date: Mon, 16 Nov 2009 15:10:00 -0500
Subject: [Dnsmasq-discuss] two stupid questions
Message-ID: <C7740C05-9337-4C67-B2F9-E1BA3F76C740@agencypja.com>

1.
The '--test' option as described in the manpage doesn't appear to do  
anything (on my Mac OS X 10.5-based box); am I doing something wrong?  
shouldn't the command be:

/usr/local/sbin/dnsmasq --test

2.
I'm unclear on how to review the logs - I see references in the  
documentation to sending SIGUSR1 and SIGUSR2 commands but don't  
understand HOW to do that.

I'd like to resolve these issues to help ensure that my machine is  
working properly before rolling it out to the office. Thanks in  
advance for your help...

-Chris


From jasiu at belsznica.pl  Mon Nov 16 20:59:02 2009
From: jasiu at belsznica.pl (Jan Psota)
Date: Mon, 16 Nov 2009 21:59:02 +0100
Subject: [Dnsmasq-discuss] Fw:  two stupid questions
Message-ID: <20091116215902.576661ce@gandalf>

Christopher Scott <cscott at agencypja.com> wrote:
> 1.
> The '--test' option as described in the manpage doesn't appear to do  
> anything (on my Mac OS X 10.5-based box); am I doing something
> wrong? shouldn't the command be:
> 
> /usr/local/sbin/dnsmasq --test
Works as expected on Gentoo Linux:

gandalf ~ # dnsmasq --test
dnsmasq: syntax check OK.
gandalf ~ # dnsmasq --test -kasj

dnsmasq: bad command line options: try --help

> 2.
> I'm unclear on how to review the logs - I see references in the  
> documentation to sending SIGUSR1 and SIGUSR2 commands but don't  
> understand HOW to do that.
If Mac OS X is as close to UNIX as I expect, running:
	killall -USR1 dnsmasq
as privileged user should provoke dnsmasq to write to log
(/var/log/messages for us) something like:

Nov 16 21:21:21 gandalf dnsmasq[5553]: time 1258402881
Nov 16 21:21:21 gandalf dnsmasq[5553]: cache size 150, 0/0 cache insertions re-used unexpired cache entries.
Nov 16 21:21:21 gandalf dnsmasq[5553]: queries forwarded 0, queries answered locally 0
Nov 16 21:21:21 gandalf dnsmasq[5553]: server 192.168.7.5#53: queries sent 0, retried or failed 0
Nov 16 21:21:21 gandalf dnsmasq[5553]: server 192.168.1.1#53: queries sent 0, retried or failed 0

Alternatively you can check dnsmasq's PID with 'ps' command and then
use
	kill -USR1 [PID]
like:
gandalf ~ # ps -C dnsmasq
  PID TTY          TIME CMD
 6136 ?        00:00:00 dnsmasq
gandalf ~ # kill -USR1 6136
gandalf ~ #
#[take a look on logs]

-- 
Jan Psota


From knud.skrald at guldberg.info  Mon Nov 16 21:59:10 2009
From: knud.skrald at guldberg.info (Knud)
Date: Mon, 16 Nov 2009 22:59:10 +0100
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4B0026EA.60002@guldberg.info>
References: <4AFF3E7B.302@guldberg.info>
	<4AFFF1B2.8050505@thekelleys.org.uk>	<4B000227.9010205@guldberg.info>	<4B001F29.8010806@thekelleys.org.uk>
	<4B0026EA.60002@guldberg.info>
Message-ID: <4B01CB2E.20902@guldberg.info>

Hi

My ISP has answerd

"You should not use ns1.bbsyd.dk any more, use only DHCP released DNS...
  a list of 3 nameservers....."

Hmm allright it seems to work with those but it would nice to be 
informed when they change something..not everybody use a simple plain 
setup !

And why have server running which do not work correctly ?

I will see if is stable in the future

Knud



Knud wrote:
> Hi Simon
> 
> I have opened a support ticket at my ISP (you right..)
> Informed them about the strange behavoir of there NS
> Let us see what happends
> 
> Otherwise they have been very stable...it's a fiberoptic connection 
> 25/25 Mbit.
> 
> Knud
> 
> Simon Kelley wrote:
>> Knud wrote:
>>> Hi Simon
>>>
>>> Just tested again...same issue again
>>>
>>> asking my upstream nameserver...
>>> camping.dba.dk => unknown
>>> www.dba.dk => ok
>>> again camping.dba.dk => now ok !
>> Yes, I tried again and saw the same thing.
>>
>> ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
>>
>> ;; QUESTION SECTION:
>> ;camping.dba.dk.			IN	A
>>
>> ;; ANSWER SECTION:
>> camping.dba.dk.		600	IN	CNAME	www.dba.dk.
>>
>> ;; Query time: 77 msec
>>
>>
>> repeat:
>>
>> ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 4, ADDITIONAL: 4
>>
>> ;; QUESTION SECTION:
>> ;camping.dba.dk.			IN	A
>>
>> ;; ANSWER SECTION:
>> camping.dba.dk.		585	IN	CNAME	www.dba.dk.
>> www.dba.dk.		1787	IN	A	91.213.17.20
>>
>> ;; AUTHORITY SECTION:
>> dba.dk.			21587	IN	NS	ns6.netnames.net.
>> dba.dk.			21587	IN	NS	ns1.netnames.net.
>> dba.dk.			21587	IN	NS	ns5.netnames.net.
>> dba.dk.			21587	IN	NS	ns2.netnames.net.
>>
>> ;; ADDITIONAL SECTION:
>> ns6.netnames.net.	168883	IN	A	199.7.69.253
>> ns1.netnames.net.	21747	IN	A	204.74.108.253
>> ns5.netnames.net.	168883	IN	A	199.7.68.253
>> ns2.netnames.net.	168883	IN	A	62.128.146.40
>>
>> ;; Query time: 46 msec
>>
>>
>> Looks like once the upstream server has the CNAME cached, it works OK,
>> but not before. There are more clues there, as it tells us the
>> authoritative nameservers (Ie the ones that hold the configured data)
>> are a netnames.net.
>>
>> We can ask one of those direct:
>>
>> srk at spike$ dig @ns6.netnames.net camping.dba.dk
>>
>> ; <<>> DiG 9.5.1-P3 <<>> @ns6.netnames.net camping.dba.dk
>> ; (1 server found)
>> ;; global options:  printcmd
>> ;; Got answer:
>> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 10510
>> ;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
>> ;; WARNING: recursion requested but not available
>>
>> ;; QUESTION SECTION:
>> ;camping.dba.dk.			IN	A
>>
>> ;; ANSWER SECTION:
>> camping.dba.dk.		600	IN	CNAME	www.dba.dk.
>>
>> ;; Query time: 22 msec
>> ;; SERVER: 199.7.69.253#53(199.7.69.253)
>> ;; WHEN: Sun Nov 15 15:20:05 2009
>> ;; MSG SIZE  rcvd: 50
>>
>>
>> Which probably shows a real problem. That nameserver always gives that
>> data without the A record.
>>
>>
>> But even more strange, the netnames.net nameservers are not the real
>> authoritative ones for dba.dk
>>
>> srk at spike:/mnt/onetouch/simon/torrents$ dig ns dba.dk
>>
>> ; <<>> DiG 9.5.1-P3 <<>> ns dba.dk
>> ;; global options:  printcmd
>> ;; Got answer:
>> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 46594
>> ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 2
>>
>> ;; QUESTION SECTION:
>> ;dba.dk.				IN	NS
>>
>> ;; ANSWER SECTION:
>> dba.dk.			43888	IN	NS	ns1.danhost.dk.
>> dba.dk.			43888	IN	NS	ns0.danhost.dk.
>>
>> ;; ADDITIONAL SECTION:
>> ns0.danhost.dk.		14276	IN	A	217.195.182.61
>> ns1.danhost.dk.		13926	IN	A	195.140.132.26
>>
>> ;; Query time: 9 msec
>> ;; SERVER: 192.168.0.4#53(192.168.0.4)
>> ;; WHEN: Sun Nov 15 15:27:37 2009
>> ;; MSG SIZE  rcvd: 100
>>
>> so the 89.184.128.193 nameserver you are using is very confused, I
>> think. It's using the wrong authoritative nameservers for dba.dk and
>> getting the wrong answers, which it is passing on. At this point it's
>> probably best to pass the whole mess over to the admins of
>> 89.184.128.193 (your ISP?) CC: me on the mail if you wish, in case I can
>> provide more information.
>>
>>> restart my dnsmasq makes it to known www.dba.dk and camping.dba.dk
>>>
>>> I am running another dnsmasq server (at a school) with another upstream
>>> nameserver...there seems no problems
>> I see the same: there's something wrong with 89.184.128.193, for sure.
>>
>>> It is another version of dnsmasq included in the FREESCO packged.
>>> But it seems not to be a dnsmasq issue but further upstream problem
>>>
>>> Just strange that this specific host gives problems and not others (to
>>> my knowlegde)
>>
>> Cheers,
>>
>> Simon.
>>
> 
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 



From simon at thekelleys.org.uk  Tue Nov 17 13:37:45 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Tue, 17 Nov 2009 13:37:45 +0000
Subject: [Dnsmasq-discuss] can resolve 1 specific host
In-Reply-To: <4B01CB2E.20902@guldberg.info>
References: <4AFF3E7B.302@guldberg.info>
	<4AFFF1B2.8050505@thekelleys.org.uk>	<4B000227.9010205@guldberg.info>	<4B001F29.8010806@thekelleys.org.uk>
	<4B0026EA.60002@guldberg.info> <4B01CB2E.20902@guldberg.info>
Message-ID: <4B02A729.5040403@thekelleys.org.uk>

Knud wrote:
> Hi
> 
> My ISP has answerd
> 
> "You should not use ns1.bbsyd.dk any more, use only DHCP released DNS...
>  a list of 3 nameservers....."
> 
> Hmm allright it seems to work with those but it would nice to be 
> informed when they change something..not everybody use a simple plain 
> setup !
> 
> And why have server running which do not work correctly ?
> 
> I will see if is stable in the future
> 

I think I'd be able to forgive my ISP such crimes if they provided 
25M/25M connectivity!


Cheers,

Simon.


From moparisthebest at hotmail.com  Tue Nov 17 18:15:43 2009
From: moparisthebest at hotmail.com (Travis)
Date: Tue, 17 Nov 2009 13:15:43 -0500
Subject: [Dnsmasq-discuss] dnsmasq appending .?
Message-ID: <c127737d0911171015x53cd696fx57e94dda89a74d77@mail.gmail.com>

Hello,

I'm having just one problem with dnsmasq. I specify addresses like
this, for example, 'address=/mediacenter/192.168.1.20', but a lookup
for 'mediacenter' fails, where a lookup for 'mediacenter.' succeeds.
I am running ubuntu server hardy, with the dnsmasq from the repos
installed:

$ dnsmasq -v
Dnsmasq version 2.41  Copyright (C) 2000-2008 Simon Kelley
Compile time options IPv6 GNU-getopt no-ISC-leasefile DBus I18N TFTP

Here are the contents of /etc/dnsmasq.conf:

domain-needed
bogus-priv
no-resolv
no-poll
server=208.67.222.222
server=208.67.220.220
address=/mytorrentflux1/192.168.1.3
interface=eth1
bind-interfaces
no-hosts
bogus-nxdomain=208.69.36.132

And a few outputs from my other computer using this server as a
caching DNS server:

mopar at killer-linux:~$ dig mediacenter

; <<>> DiG 9.5.1-P2 <<>> mediacenter
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 64254
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;mediacenter.                   IN      A

;; ANSWER SECTION:
mediacenter.            0       IN      A       192.168.1.20

;; Query time: 11 msec
;; SERVER: 192.168.1.3#53(192.168.1.3)
;; WHEN: Tue Nov 17 13:06:25 2009
;; MSG SIZE  rcvd: 45

mopar at killer-linux:~$ nslookup mediacenter
Server:         192.168.1.3
Address:        192.168.1.3#53

Name:   mediacenter.travis.com
Address: 192.168.1.1

mopar at killer-linux:~$ nslookup mediacenter.
Server:         192.168.1.3
Address:        192.168.1.3#53

Name:   mediacenter
Address: 192.168.1.20

mopar at killer-linux:~$ host mediacenter
mediacenter.travis.com has address 192.168.1.1
mopar at killer-linux:~$ host mediacenter.
mediacenter has address 192.168.1.20

As you can see, dig appears to append a . to the request
automatically, but nothing else does.  Is there a way I can get
dnsmasq to return the specified address if I don't append a . to the
request?

Thanks very much,
Travis


From ejnersan at gmail.com  Wed Nov 18 20:11:21 2009
From: ejnersan at gmail.com (Ejner Fergo)
Date: Wed, 18 Nov 2009 21:11:21 +0100
Subject: [Dnsmasq-discuss] Problem listening on only one interface
Message-ID: <ab8d3ebb0911181211h57d1bcb7lafb994143b9c626c@mail.gmail.com>

Hola all,

I'm a happy dnsmasq user for a couple of years now, though it is first
now I turn to this list to hopefully get some help.

I run dnsmasq on a server with 6 NICs, where 'eth0' talks to the
outside world and 'eth1' is for the local network. The last 4 is
bonded into 'bond0'.
My problem is I only want to have 'eth1' to listen/respond to DNS/DHCP
requests, but no matter which option I use (interface,
except-interface, listen-address) the setup breaks and the
workstations can't get an address. If all these options are commented
out, everything works just fine but some workstations see 'bond0' as
the nameserver instead of 'eth1'.

Looking at /var/log/messages on the server shows that it is random
whether 'eth1' or 'bond0' is used, for example:

Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPDISCOVER(eth1) 08:00:27:b4:b4:dd
Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPOFFER(eth1) 192.168.3.31
08:00:27:b4:b4:dd
Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPDISCOVER(bond0) 08:00:27:b4:b4:dd
Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPOFFER(bond0) 192.168.3.31
08:00:27:b4:b4:dd
Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPREQUEST(eth1)
192.168.3.31 08:00:27:b4:b4:dd
Nov 18 19:34:37 sol dnsmasq-dhcp[26878]: DHCPACK(eth1) 192.168.3.31
08:00:27:b4:b4:dd puck
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPDISCOVER(bond0) 08:00:27:b4:b4:dd
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPOFFER(bond0) 192.168.3.31
08:00:27:b4:b4:dd
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPDISCOVER(eth1) 08:00:27:b4:b4:dd
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPOFFER(eth1) 192.168.3.31
08:00:27:b4:b4:dd
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPREQUEST(bond0)
192.168.3.31 08:00:27:b4:b4:dd
Nov 18 19:34:41 sol dnsmasq-dhcp[26878]: DHCPACK(bond0) 192.168.3.31
08:00:27:b4:b4:dd puck

or

Nov 18 19:46:45 sol dnsmasq-dhcp[26878]: DHCPREQUEST(eth1)
192.168.3.30 00:0b:cd:3e:41:de
Nov 18 19:46:45 sol dnsmasq-dhcp[26878]: DHCPACK(eth1) 192.168.3.30
00:0b:cd:3e:41:de makemake
Nov 18 19:46:45 sol dnsmasq-dhcp[26878]: DHCPREQUEST(bond0)
192.168.3.30 00:0b:cd:3e:41:de
Nov 18 19:46:45 sol dnsmasq-dhcp[26878]: DHCPACK(bond0) 192.168.3.30
00:0b:cd:3e:41:de makemake

So even if it looks like 'eth1' has acknowledged the request it is
'bond0' that "won".

If I use the interface=, except-interface= or listen-address= options,
like for example:

except-interface=bond0

the messages look like this:

Nov 18 18:51:07 sol dnsmasq-dhcp[26159]: DHCPDISCOVER(eth1) 00:1e:58:a1:6e:b1
Nov 18 18:51:07 sol dnsmasq-dhcp[26159]: DHCPOFFER(eth1) 192.168.3.5
00:1e:58:a1:6e:b1
Nov 18 18:51:11 sol dnsmasq-dhcp[26159]: DHCPDISCOVER(eth1) 00:1e:58:a1:6e:b1
Nov 18 18:51:11 sol dnsmasq-dhcp[26159]: DHCPOFFER(eth1) 192.168.3.5
00:1e:58:a1:6e:b1
Nov 18 18:51:19 sol dnsmasq-dhcp[26159]: DHCPDISCOVER(eth1) 00:1e:58:a1:6e:b1
Nov 18 18:51:19 sol dnsmasq-dhcp[26159]: DHCPOFFER(eth1) 192.168.3.5
00:1e:58:a1:6e:b1
Nov 18 18:51:34 sol dnsmasq-dhcp[26159]: DHCPDISCOVER(eth1) 00:1e:58:a1:6e:b1
Nov 18 18:51:34 sol dnsmasq-dhcp[26159]: DHCPOFFER(eth1) 192.168.3.5
00:1e:58:a1:6e:b1
...
...
and so on.

No matter which options I use, with or without "bind-interfaces", I
can't get it to work for only one interface. So until now I just
didn't use these options and the workstations uses a mix of either
'eth1' or 'bond0' as nameserver, but I really would like it to be more
consistent. My dnsmasq.conf is very simple (which I like) and I read
through it some times now but can't figure out why it won't work for
only one interface?

The server is running Gentoo Linux 2.6.27 and dnsmasq is v2.50.

Thank you in advance and please let me know if the info is insufficient.

Kind regards,
Ejner Fergo


From simon at thekelleys.org.uk  Wed Nov 18 20:46:33 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 18 Nov 2009 20:46:33 +0000
Subject: [Dnsmasq-discuss] Problem listening on only one interface
In-Reply-To: <ab8d3ebb0911181211h57d1bcb7lafb994143b9c626c@mail.gmail.com>
References: <ab8d3ebb0911181211h57d1bcb7lafb994143b9c626c@mail.gmail.com>
Message-ID: <4B045D29.6080606@thekelleys.org.uk>

Ejner Fergo wrote:
> Hola all,
> 
> I'm a happy dnsmasq user for a couple of years now, though it is first
> now I turn to this list to hopefully get some help.
> 
> I run dnsmasq on a server with 6 NICs, where 'eth0' talks to the
> outside world and 'eth1' is for the local network. The last 4 is
> bonded into 'bond0'.
> My problem is I only want to have 'eth1' to listen/respond to DNS/DHCP
> requests, but no matter which option I use (interface,
> except-interface, listen-address) the setup breaks and the
> workstations can't get an address. If all these options are commented
> out, everything works just fine but some workstations see 'bond0' as
> the nameserver instead of 'eth1'.
> 
> Looking at /var/log/messages on the server shows that it is random
> whether 'eth1' or 'bond0' is used, for example:
> 


Hi,

Could you send the output from these two commands, please?

cat /proc/net/bonding/bond0
ip addr show

Cheers,

Simon.


From ejnersan at gmail.com  Thu Nov 19 14:47:00 2009
From: ejnersan at gmail.com (Ejner Fergo)
Date: Thu, 19 Nov 2009 15:47:00 +0100
Subject: [Dnsmasq-discuss] Problem listening on only one interface
In-Reply-To: <4B045D29.6080606@thekelleys.org.uk>
References: <ab8d3ebb0911181211h57d1bcb7lafb994143b9c626c@mail.gmail.com>
	<4B045D29.6080606@thekelleys.org.uk>
Message-ID: <ab8d3ebb0911190647n7cfbfeedgbcfcd10a38a97b8e@mail.gmail.com>

Hi Simon,

Thank you for replying!

Here's the output of /proc/net/bonding/bond0:

Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 0
Down Delay (ms): 0

802.3ad info
LACP rate: slow
Active Aggregator Info:
	Aggregator ID: 2
	Number of ports: 4
	Actor Key: 9
	Partner Key: 3
	Partner Mac Address: 00:1f:28:63:4c:00

Slave Interface: eth2
MII Status: up
Link Failure Count: 1
Permanent HW addr: 00:15:17:76:08:f1
Aggregator ID: 2

Slave Interface: eth3
MII Status: up
Link Failure Count: 1
Permanent HW addr: 00:15:17:76:08:f0
Aggregator ID: 2

Slave Interface: eth4
MII Status: up
Link Failure Count: 1
Permanent HW addr: 00:15:17:76:08:f3
Aggregator ID: 2

Slave Interface: eth5
MII Status: up
Link Failure Count: 1
Permanent HW addr: 00:15:17:76:08:f2
Aggregator ID: 2


And here's the output of 'ip addr show':

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
state UP qlen 1000
    link/ether 00:e0:81:b0:70:9a brd ff:ff:ff:ff:ff:ff
    inet 10.0.0.251/24 brd 10.0.0.255 scope global eth0
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc pfifo_fast
state UP qlen 1000
    link/ether 00:e0:81:b0:70:9b brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.1/24 brd 192.168.3.255 scope global eth1
4: eth2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
pfifo_fast master bond0 state UP qlen 1000
    link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
5: eth3: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
pfifo_fast master bond0 state UP qlen 1000
    link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
6: eth4: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
pfifo_fast master bond0 state UP qlen 1000
    link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
7: eth5: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
pfifo_fast master bond0 state UP qlen 1000
    link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
8: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 9000 qdisc
noqueue state UP
    link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.3.3/24 brd 192.168.3.255 scope global bond0
9: tun0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc
pfifo_fast state UNKNOWN qlen 100
    link/[65534]
    inet 10.8.0.1 peer 10.8.0.2/32 scope global tun0


Best regards,
Ejner Fergo


On Wed, Nov 18, 2009 at 9:46 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> Ejner Fergo wrote:
>> Hola all,
>>
>> I'm a happy dnsmasq user for a couple of years now, though it is first
>> now I turn to this list to hopefully get some help.
>>
>> I run dnsmasq on a server with 6 NICs, where 'eth0' talks to the
>> outside world and 'eth1' is for the local network. The last 4 is
>> bonded into 'bond0'.
>> My problem is I only want to have 'eth1' to listen/respond to DNS/DHCP
>> requests, but no matter which option I use (interface,
>> except-interface, listen-address) the setup breaks and the
>> workstations can't get an address. If all these options are commented
>> out, everything works just fine but some workstations see 'bond0' as
>> the nameserver instead of 'eth1'.
>>
>> Looking at /var/log/messages on the server shows that it is random
>> whether 'eth1' or 'bond0' is used, for example:
>>
>
>
> Hi,
>
> Could you send the output from these two commands, please?
>
> cat /proc/net/bonding/bond0
> ip addr show
>
> Cheers,
>
> Simon.
>


From simon at thekelleys.org.uk  Thu Nov 19 14:56:50 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Thu, 19 Nov 2009 14:56:50 +0000
Subject: [Dnsmasq-discuss] Problem listening on only one interface
In-Reply-To: <ab8d3ebb0911190647n7cfbfeedgbcfcd10a38a97b8e@mail.gmail.com>
References: <ab8d3ebb0911181211h57d1bcb7lafb994143b9c626c@mail.gmail.com>	
	<4B045D29.6080606@thekelleys.org.uk>
	<ab8d3ebb0911190647n7cfbfeedgbcfcd10a38a97b8e@mail.gmail.com>
Message-ID: <4B055CB2.6070507@thekelleys.org.uk>

I'm not quite clear what is going on here, but there's one thing which 
really doesn't look right and may well be confusing things: both eth1 
and bond0 are on the same subnet. If it's possible to fix that by 
changing the address of one of tham, that would be a worthwhile test.


Cheers,

Simon.


Ejner Fergo wrote:
> Hi Simon,
> 
> Thank you for replying!
> 
> Here's the output of /proc/net/bonding/bond0:
> 
> Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008)
> 
> Bonding Mode: IEEE 802.3ad Dynamic link aggregation
> Transmit Hash Policy: layer2 (0)
> MII Status: up
> MII Polling Interval (ms): 100
> Up Delay (ms): 0
> Down Delay (ms): 0
> 
> 802.3ad info
> LACP rate: slow
> Active Aggregator Info:
> 	Aggregator ID: 2
> 	Number of ports: 4
> 	Actor Key: 9
> 	Partner Key: 3
> 	Partner Mac Address: 00:1f:28:63:4c:00
> 
> Slave Interface: eth2
> MII Status: up
> Link Failure Count: 1
> Permanent HW addr: 00:15:17:76:08:f1
> Aggregator ID: 2
> 
> Slave Interface: eth3
> MII Status: up
> Link Failure Count: 1
> Permanent HW addr: 00:15:17:76:08:f0
> Aggregator ID: 2
> 
> Slave Interface: eth4
> MII Status: up
> Link Failure Count: 1
> Permanent HW addr: 00:15:17:76:08:f3
> Aggregator ID: 2
> 
> Slave Interface: eth5
> MII Status: up
> Link Failure Count: 1
> Permanent HW addr: 00:15:17:76:08:f2
> Aggregator ID: 2
> 
> 
> And here's the output of 'ip addr show':
> 
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>     inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
> state UP qlen 1000
>     link/ether 00:e0:81:b0:70:9a brd ff:ff:ff:ff:ff:ff
>     inet 10.0.0.251/24 brd 10.0.0.255 scope global eth0
> 3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc pfifo_fast
> state UP qlen 1000
>     link/ether 00:e0:81:b0:70:9b brd ff:ff:ff:ff:ff:ff
>     inet 192.168.3.1/24 brd 192.168.3.255 scope global eth1
> 4: eth2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
> pfifo_fast master bond0 state UP qlen 1000
>     link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
> 5: eth3: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
> pfifo_fast master bond0 state UP qlen 1000
>     link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
> 6: eth4: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
> pfifo_fast master bond0 state UP qlen 1000
>     link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
> 7: eth5: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 9000 qdisc
> pfifo_fast master bond0 state UP qlen 1000
>     link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
> 8: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 9000 qdisc
> noqueue state UP
>     link/ether 00:15:17:76:08:f1 brd ff:ff:ff:ff:ff:ff
>     inet 192.168.3.3/24 brd 192.168.3.255 scope global bond0
> 9: tun0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc
> pfifo_fast state UNKNOWN qlen 100
>     link/[65534]
>     inet 10.8.0.1 peer 10.8.0.2/32 scope global tun0
> 
> 
> Best regards,
> Ejner Fergo
> 
> 
> On Wed, Nov 18, 2009 at 9:46 PM, Simon Kelley <simon at thekelleys.org.uk> wrote:
>> Ejner Fergo wrote:
>>> Hola all,
>>>
>>> I'm a happy dnsmasq user for a couple of years now, though it is first
>>> now I turn to this list to hopefully get some help.
>>>
>>> I run dnsmasq on a server with 6 NICs, where 'eth0' talks to the
>>> outside world and 'eth1' is for the local network. The last 4 is
>>> bonded into 'bond0'.
>>> My problem is I only want to have 'eth1' to listen/respond to DNS/DHCP
>>> requests, but no matter which option I use (interface,
>>> except-interface, listen-address) the setup breaks and the
>>> workstations can't get an address. If all these options are commented
>>> out, everything works just fine but some workstations see 'bond0' as
>>> the nameserver instead of 'eth1'.
>>>
>>> Looking at /var/log/messages on the server shows that it is random
>>> whether 'eth1' or 'bond0' is used, for example:
>>>
>>
>> Hi,
>>
>> Could you send the output from these two commands, please?
>>
>> cat /proc/net/bonding/bond0
>> ip addr show
>>
>> Cheers,
>>
>> Simon.
>>
> 



From adam.ant at cyberspaceroad.com  Fri Nov 20 20:38:13 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Fri, 20 Nov 2009 20:38:13 +0000
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
Message-ID: <4B06FE35.6010701@cyberspaceroad.com>

Hi

I have a problem which is not giving up its solution to any of my google searches.

I have a lan with a gateway machine running an ADSL modem and two NICs with 
iptables and dnsmasq.

It also runs mysql and tomcat but is currently just a simple gateway, I'm not 
trying to configure any DMZ or fancier stuff like that.

My problem is that I can access mysql using 'localhost:3306' but I can't access 
it on the same box when using the machine name e.g. 'isengard:3306' and my guess 
is that I have mis-configured either dnsmasq or iptables.

I figure that my command "mysql --host=isengard" is probably being resolved as 
external and then getting blocked by the firewall.

My dnsmasq config file, based on the example config but with comments removed, is:

domain-needed
bogus-priv
filterwin2k
local=/localdomain/
domain=localdomain
dhcp-range=192.168.0.3,192.168.0.254
dhcp-option=option:router,192.168.0.2
dhcp-option=option:mtu,1500

and my resolv.conf file is:

nameserver 194.74.65.68

and I think this is getting continually rewritten by dhcp with the nameserver 
info from the dhcp server on the modem which gives the outside NIC its internet 
ip address.

Does this make any sense? Or rather does anyone see where my situation is 
foobarred?


Thanks
Adam


From santiago at zarate.net.ve  Fri Nov 20 02:49:28 2009
From: santiago at zarate.net.ve (Santiago Zarate)
Date: Thu, 19 Nov 2009 22:19:28 -0430
Subject: [Dnsmasq-discuss] two stupid questions
In-Reply-To: <C7740C05-9337-4C67-B2F9-E1BA3F76C740@agencypja.com>
References: <C7740C05-9337-4C67-B2F9-E1BA3F76C740@agencypja.com>
Message-ID: <200911192219.29143.santiago@zarate.net.ve>

>1.
>The '--test' option as described in the manpage doesn't appear to do
>anything (on my Mac OS X 10.5-based box); am I doing something wrong?
>shouldn't the command be:
>
>/usr/local/sbin/dnsmasq --test

Actually the --test option works fine for me... 

santiago at braindeath:~$ dnsmasq --test
dnsmasq: sintaxis correcta.

santiago at braindeath:~$ /usr/sbin/dnsmasq --test
dnsmasq: sintaxis correcta.

also, my dnsmasq version is: 

santiago at braindeath:~$ dnsmasq -v
Dnsmasq versi?n 2.50  Copyright (C) 2000-2009 Simon Kelley
Opciones de compilaci?n IPv6 GNU-getopt DBus I18N DHCP TFTP


>2.
>I'm unclear on how to review the logs - I see references in the
>documentation to sending SIGUSR1 and SIGUSR2 commands but don't
>understand HOW to do that.

Its not exactly a command, its a signal... i use htop to send signals to some process and 
then watch the log for changes... in OSX, i have no idea... but looking on google might be 
a good idea...  i have few mac pc's at office.. if i remember... ill check later... 

>I'd like to resolve these issues to help ensure that my machine is
>working properly before rolling it out to the office. Thanks in
>advance for your help...

-- 
Santiago Zarate
santiago at zarate.net.ve
(+58) 4129864175
(+58) 4241073905


From lee.essen at me.com  Fri Nov 20 22:37:19 2009
From: lee.essen at me.com (Lee Essen)
Date: Fri, 20 Nov 2009 22:37:19 +0000
Subject: [Dnsmasq-discuss] Problem compiling 2.51 on Solaris 2009.06
Message-ID: <134545201205494196899049187214966559998-Webmail@me.com>

Hi,

I've been a dnsmasq user for quite a few years on both Solaris and Linux, but it's been a while since I've tried to build on Solaris. I've just tried to build 2.51 on the latest Solaris build and there seems to be a number of really fundamental issues.

I notice Solaris mentioned quite a few times in the changelog so there seems to be a continued effort around this, and I certainly don't remember any issues in the past, however unless I'm doing something fundamentally wrong (which is a distinct possibility) I think a few tweaks are needed...

I have attached a patch that covers all the changes I needed to make, although I must confess that whilst it's now compiles I haven't tested it yet.

Three main problems...

1.  config.h is included after sys/socket.h -- this means that _XPG4_2 isn't defined yet and therefore all of the CMSG_* defines get missed (CMSG_SPACE for example.)

2. struct ifreq seems to have become struct lifrect (at least if you need ifr_mtu), so this impacts a number of references in two or three files.

3. LOG_PRI doesn't seem to exist at all.

I'm not suggesting my patch is the right solution, it's probably far from it, but it's enough to highlight the issues.

If I'm missed something obvious then please let me know, otherwise I'll continue to test and see if I can come up with a less disruptive way of incorporating these fixes.

Thanks,

Lee. 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-2.51-solaris.patch
Type: application/octet-stream
Size: 4473 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091120/5a23f33c/attachment.obj 

From simon at thekelleys.org.uk  Sat Nov 21 09:30:19 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Sat, 21 Nov 2009 09:30:19 +0000
Subject: [Dnsmasq-discuss] Problem compiling 2.51 on Solaris 2009.06
In-Reply-To: <134545201205494196899049187214966559998-Webmail@me.com>
References: <134545201205494196899049187214966559998-Webmail@me.com>
Message-ID: <4B07B32B.9040503@thekelleys.org.uk>

Lee Essen wrote:
> Hi,
> 
> I've been a dnsmasq user for quite a few years on both Solaris and
> Linux, but it's been a while since I've tried to build on Solaris.
> I've just tried to build 2.51 on the latest Solaris build and there
> seems to be a number of really fundamental issues.
> 
> I notice Solaris mentioned quite a few times in the changelog so
> there seems to be a continued effort around this, and I certainly
> don't remember any issues in the past, however unless I'm doing
> something fundamentally wrong (which is a distinct possibility) I
> think a few tweaks are needed...
> 
> I have attached a patch that covers all the changes I needed to make,
> although I must confess that whilst it's now compiles I haven't
> tested it yet.
> 
> Three main problems...
> 
> 1.  config.h is included after sys/socket.h -- this means that
> _XPG4_2 isn't defined yet and therefore all of the CMSG_* defines get
> missed (CMSG_SPACE for example.)
> 
> 2. struct ifreq seems to have become struct lifrect (at least if you
> need ifr_mtu), so this impacts a number of references in two or three
> files.
> 
> 3. LOG_PRI doesn't seem to exist at all.
> 
> I'm not suggesting my patch is the right solution, it's probably far
> from it, but it's enough to highlight the issues.
> 
> If I'm missed something obvious then please let me know, otherwise
> I'll continue to test and see if I can come up with a less disruptive
> way of incorporating these fixes.
> 
Many thanks for this.

By some weird coincidence, I was working on this yesterday too. I found 
and fixed your problems 1 and 3 but didn't get bitten by 2. Does that 
turn up as a compile-time problem, or just run-time?

Anyway, the results of my efforts are at

http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test7.tar.gz

which compile, but have not been functionally tested yet.

Please could you check that and let me know if it's OK, or if I need to 
work-in your lifr stuff?


Cheers,

Simon.


From lee.essen at me.com  Sat Nov 21 11:59:10 2009
From: lee.essen at me.com (Lee Essen)
Date: Sat, 21 Nov 2009 11:59:10 +0000
Subject: [Dnsmasq-discuss] Problem compiling 2.51 on Solaris 2009.06
In-Reply-To: <88975658192063584747046041514764610356-Webmail@me.com>
References: <88975658192063584747046041514764610356-Webmail@me.com>
Message-ID: <116875675832733763473130959593417804803-Webmail@me.com>

 
On Saturday, November 21, 2009, at 09:30AM, "Simon Kelley" <simon at thekelleys.org.uk> wrote:
>Lee Essen wrote:
>> Hi,
>> 
>> I've been a dnsmasq user for quite a few years on both Solaris and
>> Linux, but it's been a while since I've tried to build on Solaris.
>> I've just tried to build 2.51 on the latest Solaris build and there
>> seems to be a number of really fundamental issues.
>> 
>> I notice Solaris mentioned quite a few times in the changelog so
>> there seems to be a continued effort around this, and I certainly
>> don't remember any issues in the past, however unless I'm doing
>> something fundamentally wrong (which is a distinct possibility) I
>> think a few tweaks are needed...
>> 
>> I have attached a patch that covers all the changes I needed to make,
>> although I must confess that whilst it's now compiles I haven't
>> tested it yet.
>> 
>> Three main problems...
>> 
>> 1.  config.h is included after sys/socket.h -- this means that
>> _XPG4_2 isn't defined yet and therefore all of the CMSG_* defines get
>> missed (CMSG_SPACE for example.)
>> 
>> 2. struct ifreq seems to have become struct lifrect (at least if you
>> need ifr_mtu), so this impacts a number of references in two or three
>> files.
>> 
>> 3. LOG_PRI doesn't seem to exist at all.
>> 
>> I'm not suggesting my patch is the right solution, it's probably far
>> from it, but it's enough to highlight the issues.
>> 
>> If I'm missed something obvious then please let me know, otherwise
>> I'll continue to test and see if I can come up with a less disruptive
>> way of incorporating these fixes.
>> 
>Many thanks for this.
>
>By some weird coincidence, I was working on this yesterday too. I found 
>and fixed your problems 1 and 3 but didn't get bitten by 2. Does that 
>turn up as a compile-time problem, or just run-time?
>
>Anyway, the results of my efforts are at
>
>http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test7.tar.gz
>
>which compile, but have not been functionally tested yet.
>
>Please could you check that and let me know if it's OK, or if I need to 
>work-in your lifr stuff?
>
>
>Cheers,
>
>Simon.
>

Simon,

I still see the same problem with the ifr bits - this is on the latest Solaris release ... it looks like it's just the ifr_mtu field that's the problem, according to the if.h file the struct ifreq is kept for compatibility but there is no ifr_mtu field as part of the union.

I've attached if.h for your info, just in case you have an older version.

make[1]: Entering directory `/site/internal/build/dnsmasq-2.52test7/src'
cc -Wall -W -O2     -c cache.c
cc -Wall -W -O2     -c rfc1035.c
cc -Wall -W -O2     -c util.c
cc -Wall -W -O2     -c option.c
cc -Wall -W -O2     -c forward.c
cc -Wall -W -O2     -c network.c
network.c: In function `iface_allowed':
network.c:149: error: structure has no member named `ifr_mtu'
network.c: At top level:
network.c:555: warning: unused parameter 'intname'
network.c:39: warning: unused parameter 'fd'
make[1]: *** [network.o] Error 1
make[1]: Leaving directory `/site/internal/build/dnsmasq-2.52test7/src'
make: *** [dnsmasq] Error 2

Regards,

Lee.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: if.h
Type: application/octet-stream
Size: 26310 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091121/f184b337/attachment-0001.obj 

From simon at thekelleys.org.uk  Sat Nov 21 12:49:44 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Sat, 21 Nov 2009 12:49:44 +0000
Subject: [Dnsmasq-discuss] Problem compiling 2.51 on Solaris 2009.06
In-Reply-To: <116875675832733763473130959593417804803-Webmail@me.com>
References: <88975658192063584747046041514764610356-Webmail@me.com>
	<116875675832733763473130959593417804803-Webmail@me.com>
Message-ID: <4B07E1E8.90106@thekelleys.org.uk>

Lee Essen wrote:
>  
> On Saturday, November 21, 2009, at 09:30AM, "Simon Kelley" <simon at thekelleys.org.uk> wrote:
>> Lee Essen wrote:
>>> Hi,
>>>
>>> I've been a dnsmasq user for quite a few years on both Solaris and
>>> Linux, but it's been a while since I've tried to build on Solaris.
>>> I've just tried to build 2.51 on the latest Solaris build and there
>>> seems to be a number of really fundamental issues.
>>>
>>> I notice Solaris mentioned quite a few times in the changelog so
>>> there seems to be a continued effort around this, and I certainly
>>> don't remember any issues in the past, however unless I'm doing
>>> something fundamentally wrong (which is a distinct possibility) I
>>> think a few tweaks are needed...
>>>
>>> I have attached a patch that covers all the changes I needed to make,
>>> although I must confess that whilst it's now compiles I haven't
>>> tested it yet.
>>>
>>> Three main problems...
>>>
>>> 1.  config.h is included after sys/socket.h -- this means that
>>> _XPG4_2 isn't defined yet and therefore all of the CMSG_* defines get
>>> missed (CMSG_SPACE for example.)
>>>
>>> 2. struct ifreq seems to have become struct lifrect (at least if you
>>> need ifr_mtu), so this impacts a number of references in two or three
>>> files.
>>>
>>> 3. LOG_PRI doesn't seem to exist at all.
>>>
>>> I'm not suggesting my patch is the right solution, it's probably far
>>> from it, but it's enough to highlight the issues.
>>>
>>> If I'm missed something obvious then please let me know, otherwise
>>> I'll continue to test and see if I can come up with a less disruptive
>>> way of incorporating these fixes.
>>>
>> Many thanks for this.
>>
>> By some weird coincidence, I was working on this yesterday too. I found 
>> and fixed your problems 1 and 3 but didn't get bitten by 2. Does that 
>> turn up as a compile-time problem, or just run-time?
>>
>> Anyway, the results of my efforts are at
>>
>> http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test7.tar.gz
>>
>> which compile, but have not been functionally tested yet.
>>
>> Please could you check that and let me know if it's OK, or if I need to 
>> work-in your lifr stuff?
>>
>>
>> Cheers,
>>
>> Simon.
>>
> 
> Simon,
> 
> I still see the same problem with the ifr bits - this is on the latest Solaris release ... it looks like it's just the ifr_mtu field that's the problem, according to the if.h file the struct ifreq is kept for compatibility but there is no ifr_mtu field as part of the union.
> 
> I've attached if.h for your info, just in case you have an older version.
> 
> make[1]: Entering directory `/site/internal/build/dnsmasq-2.52test7/src'
> cc -Wall -W -O2     -c cache.c
> cc -Wall -W -O2     -c rfc1035.c
> cc -Wall -W -O2     -c util.c
> cc -Wall -W -O2     -c option.c
> cc -Wall -W -O2     -c forward.c
> cc -Wall -W -O2     -c network.c
> network.c: In function `iface_allowed':
> network.c:149: error: structure has no member named `ifr_mtu'
> network.c: At top level:
> network.c:555: warning: unused parameter 'intname'
> network.c:39: warning: unused parameter 'fd'
> make[1]: *** [network.o] Error 1
> make[1]: Leaving directory `/site/internal/build/dnsmasq-2.52test7/src'
> make: *** [dnsmasq] Error 2
> 
> Regards,
> 
> Lee.


OK, it looks like that's a bug in Solaris 2009.06, since fixed in the
svn_127 version I was using. Rather than bending lots of code out of
shape to fix a transient problem, the following hack should work:

--- src/dnsmasq.h.old	2009-11-21 12:44:49.000000000 +0000
+++ src/dnsmasq.h	2009-11-21 12:48:46.000000000 +0000
@@ -72,6 +72,10 @@
 #include <sys/un.h>
 #include <limits.h>
 #include <net/if.h>
+#if defined(HAVE_SOLARIS_NETWORK) && !defined(ifr_mtu)
+/* Some solaris net/if./h omit this. */
+#  define ifr_mtu  ifr_ifru.ifru_metric
+#endif
 #include <unistd.h>
 #include <stdio.h>
 #include <string.h>

Please could you verify that?

Cheers,

Simon.



From lee.essen at me.com  Sat Nov 21 14:16:37 2009
From: lee.essen at me.com (Lee Essen)
Date: Sat, 21 Nov 2009 14:16:37 +0000
Subject: [Dnsmasq-discuss] Problem compiling 2.51 on Solaris 2009.06
In-Reply-To: <81633087582746095944215240943366106172-Webmail@me.com>
References: <81633087582746095944215240943366106172-Webmail@me.com>
Message-ID: <166071646768291866683862012899278593898-Webmail@me.com>

 
On Saturday, November 21, 2009, at 12:49PM, "Simon Kelley" <simon at thekelleys.org.uk> wrote:
>Lee Essen wrote:
>>  
>> On Saturday, November 21, 2009, at 09:30AM, "Simon Kelley" <simon at thekelleys.org.uk> wrote:
>>> Lee Essen wrote:
>>>> Hi,
>>>>
>>>> I've been a dnsmasq user for quite a few years on both Solaris and
>>>> Linux, but it's been a while since I've tried to build on Solaris.
>>>> I've just tried to build 2.51 on the latest Solaris build and there
>>>> seems to be a number of really fundamental issues.
>>>>
>>>> I notice Solaris mentioned quite a few times in the changelog so
>>>> there seems to be a continued effort around this, and I certainly
>>>> don't remember any issues in the past, however unless I'm doing
>>>> something fundamentally wrong (which is a distinct possibility) I
>>>> think a few tweaks are needed...
>>>>
>>>> I have attached a patch that covers all the changes I needed to make,
>>>> although I must confess that whilst it's now compiles I haven't
>>>> tested it yet.
>>>>
>>>> Three main problems...
>>>>
>>>> 1.  config.h is included after sys/socket.h -- this means that
>>>> _XPG4_2 isn't defined yet and therefore all of the CMSG_* defines get
>>>> missed (CMSG_SPACE for example.)
>>>>
>>>> 2. struct ifreq seems to have become struct lifrect (at least if you
>>>> need ifr_mtu), so this impacts a number of references in two or three
>>>> files.
>>>>
>>>> 3. LOG_PRI doesn't seem to exist at all.
>>>>
>>>> I'm not suggesting my patch is the right solution, it's probably far
>>>> from it, but it's enough to highlight the issues.
>>>>
>>>> If I'm missed something obvious then please let me know, otherwise
>>>> I'll continue to test and see if I can come up with a less disruptive
>>>> way of incorporating these fixes.
>>>>
>>> Many thanks for this.
>>>
>>> By some weird coincidence, I was working on this yesterday too. I found 
>>> and fixed your problems 1 and 3 but didn't get bitten by 2. Does that 
>>> turn up as a compile-time problem, or just run-time?
>>>
>>> Anyway, the results of my efforts are at
>>>
>>> http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test7.tar.gz
>>>
>>> which compile, but have not been functionally tested yet.
>>>
>>> Please could you check that and let me know if it's OK, or if I need to 
>>> work-in your lifr stuff?
>>>
>>>
>>> Cheers,
>>>
>>> Simon.
>>>
>> 
>> Simon,
>> 
>> I still see the same problem with the ifr bits - this is on the latest Solaris release ... it looks like it's just the ifr_mtu field that's the problem, according to the if.h file the struct ifreq is kept for compatibility but there is no ifr_mtu field as part of the union.
>> 
>> I've attached if.h for your info, just in case you have an older version.
>> 
>> make[1]: Entering directory `/site/internal/build/dnsmasq-2.52test7/src'
>> cc -Wall -W -O2     -c cache.c
>> cc -Wall -W -O2     -c rfc1035.c
>> cc -Wall -W -O2     -c util.c
>> cc -Wall -W -O2     -c option.c
>> cc -Wall -W -O2     -c forward.c
>> cc -Wall -W -O2     -c network.c
>> network.c: In function `iface_allowed':
>> network.c:149: error: structure has no member named `ifr_mtu'
>> network.c: At top level:
>> network.c:555: warning: unused parameter 'intname'
>> network.c:39: warning: unused parameter 'fd'
>> make[1]: *** [network.o] Error 1
>> make[1]: Leaving directory `/site/internal/build/dnsmasq-2.52test7/src'
>> make: *** [dnsmasq] Error 2
>> 
>> Regards,
>> 
>> Lee.
>
>
>OK, it looks like that's a bug in Solaris 2009.06, since fixed in the
>svn_127 version I was using. Rather than bending lots of code out of
>shape to fix a transient problem, the following hack should work:
>
>--- src/dnsmasq.h.old	2009-11-21 12:44:49.000000000 +0000
>+++ src/dnsmasq.h	2009-11-21 12:48:46.000000000 +0000
>@@ -72,6 +72,10 @@
> #include <sys/un.h>
> #include <limits.h>
> #include <net/if.h>
>+#if defined(HAVE_SOLARIS_NETWORK) && !defined(ifr_mtu)
>+/* Some solaris net/if./h omit this. */
>+#  define ifr_mtu  ifr_ifru.ifru_metric
>+#endif
> #include <unistd.h>
> #include <stdio.h>
> #include <string.h>
>
>Please could you verify that?
>
>Cheers,
>
>Simon.

Simon,

Yes this seems to work fine and is much nicer. 

I'm just puting together a quick script that will create a new Solaris ipkg with the smf stuff in it, any interest in including this in the contrib section?

Regards,

Lee.


From adam.ant at cyberspaceroad.com  Mon Nov 23 18:28:06 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Mon, 23 Nov 2009 18:28:06 +0000
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B06FE35.6010701@cyberspaceroad.com>
References: <4B06FE35.6010701@cyberspaceroad.com>
Message-ID: <4B0AD436.5090605@cyberspaceroad.com>

Thought I had a simple problem but I don't really find anything relevant on the 
web and I'm not getting any responses to my questions here.

Just a pointer in the right direction would be helpful - something to put me 
back on the scent?

Thanks
Adam

Adam Hardy on 20/11/09 20:38, wrote:
> I have a lan with a gateway machine running an ADSL modem and two NICs with 
> iptables and dnsmasq.
> 
> It also runs mysql and tomcat but is currently just a simple gateway, I'm not 
> trying to configure any DMZ or fancier stuff like that.
> 
> My problem is that I can access mysql using 'localhost:3306' but I can't access 
> it on the same box when using the machine name e.g. 'isengard:3306' and my guess 
> is that I have mis-configured either dnsmasq or iptables.
> 
> I figure that my command "mysql --host=isengard" is probably being resolved as 
> external and then getting blocked by the firewall.
> 
> My dnsmasq config file, based on the example config but with comments removed, is:
> 
> domain-needed
> bogus-priv
> filterwin2k
> local=/localdomain/
> domain=localdomain
> dhcp-range=192.168.0.3,192.168.0.254
> dhcp-option=option:router,192.168.0.2
> dhcp-option=option:mtu,1500
> 
> and my resolv.conf file is:
> 
> nameserver 194.74.65.68
> 
> and I think this is getting continually rewritten by dhcp with the nameserver 
> info from the dhcp server on the modem which gives the outside NIC its internet 
> ip address.
> 
> Does this make any sense? Or rather does anyone see where my situation is 
> foobarred?



From mark at jemms.net  Mon Nov 23 19:00:21 2009
From: mark at jemms.net (Mark Beierl)
Date: Mon, 23 Nov 2009 14:00:21 -0500
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0AD436.5090605@cyberspaceroad.com>
References: <4B06FE35.6010701@cyberspaceroad.com>
	<4B0AD436.5090605@cyberspaceroad.com>
Message-ID: <4B0ADBC5.2000104@jemms.net>

Silly thought but - is mysql configured to listen to 127.0.0.1 only? 
Something like

sudo netstat -napt | grep 3306

ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.

Regards,
Mark

Adam Hardy wrote:
> Thought I had a simple problem but I don't really find anything relevant on the 
> web and I'm not getting any responses to my questions here.
>
> Just a pointer in the right direction would be helpful - something to put me 
> back on the scent?
>
> Thanks
> Adam
>
> Adam Hardy on 20/11/09 20:38, wrote:
>   
>> I have a lan with a gateway machine running an ADSL modem and two NICs with 
>> iptables and dnsmasq.
>>
>> It also runs mysql and tomcat but is currently just a simple gateway, I'm not 
>> trying to configure any DMZ or fancier stuff like that.
>>
>> My problem is that I can access mysql using 'localhost:3306' but I can't access 
>> it on the same box when using the machine name e.g. 'isengard:3306' and my guess 
>> is that I have mis-configured either dnsmasq or iptables.
>>
>> I figure that my command "mysql --host=isengard" is probably being resolved as 
>> external and then getting blocked by the firewall.
>>
>> My dnsmasq config file, based on the example config but with comments removed, is:
>>
>> domain-needed
>> bogus-priv
>> filterwin2k
>> local=/localdomain/
>> domain=localdomain
>> dhcp-range=192.168.0.3,192.168.0.254
>> dhcp-option=option:router,192.168.0.2
>> dhcp-option=option:mtu,1500
>>
>> and my resolv.conf file is:
>>
>> nameserver 194.74.65.68
>>
>> and I think this is getting continually rewritten by dhcp with the nameserver 
>> info from the dhcp server on the modem which gives the outside NIC its internet 
>> ip address.
>>
>> Does this make any sense? Or rather does anyone see where my situation is 
>> foobarred?
>>     
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>   


From b-morgan at concentric.net  Mon Nov 23 21:24:44 2009
From: b-morgan at concentric.net (Brad Morgan)
Date: Mon, 23 Nov 2009 14:24:44 -0700
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0AD436.5090605@cyberspaceroad.com>
References: <4B06FE35.6010701@cyberspaceroad.com>
	<4B0AD436.5090605@cyberspaceroad.com>
Message-ID: <007e01ca6c83$61368900$23a39b00$@net>

> My dnsmasq config file, based on the example config but with comments
removed, is:
> 
> domain-needed
> bogus-priv
> filterwin2k
> local=/localdomain/
> domain=localdomain
> dhcp-range=192.168.0.3,192.168.0.254
> dhcp-option=option:router,192.168.0.2
> dhcp-option=option:mtu,1500
> 
> and my resolv.conf file is:
> 
> nameserver 194.74.65.68

I have a very similar configuration. I think you are close but you may need
to tweak your DHCP client. I also have a couple additional lines in my
dnsmasq.conf.

My /etc/dnsmasq.conf:

domain-needed
bogus-priv
server=/morgan.local/127.0.0.1
local=/morgan.local/
expand-hosts
domain=morgan.local
dhcp-range=192.168.0.100,192.168.0.199,4h
dhcp-option=3,192.168.0.1
dhcp-option=6,192.168.0.1
bogus-nxdomain=208.67.217.130 #OpenDNS
bogus-nxdomain=208.69.32.132 #OpenDNS
log-facility=/var/log/dnsmasq

Take a look at man dhclient.conf to see what might make sense for your
configuration. I think "prepend domain-name-servers 127.0.0.1" may be all
you need. 

My /etc/dhclient-eth0.conf:

supersede domain-name "morgan.local"; 
supersede domain-name-servers 127.0.0.1,208.67.222.222,208.67.220.220;
#OpenDNS
send host-name "bricknix";  # temporary RHL ifup addition

Regards,

Brad





From adam.ant at cyberspaceroad.com  Tue Nov 24 11:31:44 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Tue, 24 Nov 2009 11:31:44 +0000
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0ADBC5.2000104@jemms.net>
References: <4B06FE35.6010701@cyberspaceroad.com>
	<4B0AD436.5090605@cyberspaceroad.com> <4B0ADBC5.2000104@jemms.net>
Message-ID: <4B0BC420.2000203@cyberspaceroad.com>

You're right. The result from netstat was:


tcp   0   0 127.0.0.1:3306   0.0.0.0:*  LISTEN 2557/mysqld

after changing the bindaddress config in the mysql config as per the docs to 
free up networking, it then gives this result:

adam at isengard:~$ sudo netstat -napt | grep 3306
tcp   0   0 127.0.1.1:3306    0.0.0.0:*       LISTEN      16473/mysqld
tcp   0   0 127.0.1.1:53067   127.0.1.1:3306  TIME_WAIT   -

which looks weird. But then it's probably just because I don't have much 
experience in this area. Why has it switched over to 127.0.1.1 ?

Thanks
Adam


Mark Beierl on 23/11/09 19:00, wrote:
> Silly thought but - is mysql configured to listen to 127.0.0.1 only? 
> Something like
> 
> sudo netstat -napt | grep 3306
> 
> ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.
> 
> Regards,
> Mark
> 
> Adam Hardy wrote:
>> Thought I had a simple problem but I don't really find anything 
>> relevant on the web and I'm not getting any responses to my questions 
>> here.
>>
>> Just a pointer in the right direction would be helpful - something to 
>> put me back on the scent?
>>
>> Thanks
>> Adam
>>
>> Adam Hardy on 20/11/09 20:38, wrote:
>>  
>>> I have a lan with a gateway machine running an ADSL modem and two 
>>> NICs with iptables and dnsmasq.
>>>
>>> It also runs mysql and tomcat but is currently just a simple gateway, 
>>> I'm not trying to configure any DMZ or fancier stuff like that.
>>>
>>> My problem is that I can access mysql using 'localhost:3306' but I 
>>> can't access it on the same box when using the machine name e.g. 
>>> 'isengard:3306' and my guess is that I have mis-configured either 
>>> dnsmasq or iptables.
>>>
>>> I figure that my command "mysql --host=isengard" is probably being 
>>> resolved as external and then getting blocked by the firewall.
>>>
>>> My dnsmasq config file, based on the example config but with comments 
>>> removed, is:
>>>
>>> domain-needed
>>> bogus-priv
>>> filterwin2k
>>> local=/localdomain/
>>> domain=localdomain
>>> dhcp-range=192.168.0.3,192.168.0.254
>>> dhcp-option=option:router,192.168.0.2
>>> dhcp-option=option:mtu,1500
>>>
>>> and my resolv.conf file is:
>>>
>>> nameserver 194.74.65.68
>>>
>>> and I think this is getting continually rewritten by dhcp with the 
>>> nameserver info from the dhcp server on the modem which gives the 
>>> outside NIC its internet ip address.
>>>
>>> Does this make any sense? Or rather does anyone see where my 
>>> situation is foobarred?
>>>     
>>
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>   
> 



From mark at jemms.net  Tue Nov 24 14:28:48 2009
From: mark at jemms.net (Mark Beierl)
Date: Tue, 24 Nov 2009 09:28:48 -0500
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0BC420.2000203@cyberspaceroad.com>
References: <4B06FE35.6010701@cyberspaceroad.com>	<4B0AD436.5090605@cyberspaceroad.com>
	<4B0ADBC5.2000104@jemms.net> <4B0BC420.2000203@cyberspaceroad.com>
Message-ID: <4B0BEDA0.1050304@jemms.net>

The TIME_WAIT is not an active socket, it's the remnant of a previous 
connection.  I have no idea at all why mysqld has moved to 127.0.1.1.  
Is the bind address config line set to the host name and is the host 
name entry in /etc/hosts 127.0.1.1?

Unfortunately, I know very little about mysql, so I can't point you in 
the right direction for configuration...

Regards,
Mark

Adam Hardy wrote:
> You're right. The result from netstat was:
>
>
> tcp   0   0 127.0.0.1:3306   0.0.0.0:*  LISTEN 2557/mysqld
>
> after changing the bindaddress config in the mysql config as per the docs to 
> free up networking, it then gives this result:
>
> adam at isengard:~$ sudo netstat -napt | grep 3306
> tcp   0   0 127.0.1.1:3306    0.0.0.0:*       LISTEN      16473/mysqld
> tcp   0   0 127.0.1.1:53067   127.0.1.1:3306  TIME_WAIT   -
>
> which looks weird. But then it's probably just because I don't have much 
> experience in this area. Why has it switched over to 127.0.1.1 ?
>
> Thanks
> Adam
>
>
> Mark Beierl on 23/11/09 19:00, wrote:
>   
>> Silly thought but - is mysql configured to listen to 127.0.0.1 only? 
>> Something like
>>
>> sudo netstat -napt | grep 3306
>>
>> ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.
>>
>> Regards,
>> Mark
>>
>> Adam Hardy wrote:
>>     
>>> Thought I had a simple problem but I don't really find anything 
>>> relevant on the web and I'm not getting any responses to my questions 
>>> here.
>>>
>>> Just a pointer in the right direction would be helpful - something to 
>>> put me back on the scent?
>>>
>>> Thanks
>>> Adam
>>>
>>> Adam Hardy on 20/11/09 20:38, wrote:
>>>  
>>>       
>>>> I have a lan with a gateway machine running an ADSL modem and two 
>>>> NICs with iptables and dnsmasq.
>>>>
>>>> It also runs mysql and tomcat but is currently just a simple gateway, 
>>>> I'm not trying to configure any DMZ or fancier stuff like that.
>>>>
>>>> My problem is that I can access mysql using 'localhost:3306' but I 
>>>> can't access it on the same box when using the machine name e.g. 
>>>> 'isengard:3306' and my guess is that I have mis-configured either 
>>>> dnsmasq or iptables.
>>>>
>>>> I figure that my command "mysql --host=isengard" is probably being 
>>>> resolved as external and then getting blocked by the firewall.
>>>>
>>>> My dnsmasq config file, based on the example config but with comments 
>>>> removed, is:
>>>>
>>>> domain-needed
>>>> bogus-priv
>>>> filterwin2k
>>>> local=/localdomain/
>>>> domain=localdomain
>>>> dhcp-range=192.168.0.3,192.168.0.254
>>>> dhcp-option=option:router,192.168.0.2
>>>> dhcp-option=option:mtu,1500
>>>>
>>>> and my resolv.conf file is:
>>>>
>>>> nameserver 194.74.65.68
>>>>
>>>> and I think this is getting continually rewritten by dhcp with the 
>>>> nameserver info from the dhcp server on the modem which gives the 
>>>> outside NIC its internet ip address.
>>>>
>>>> Does this make any sense? Or rather does anyone see where my 
>>>> situation is foobarred?
>>>>     
>>>>         
>>> _______________________________________________
>>> Dnsmasq-discuss mailing list
>>> Dnsmasq-discuss at lists.thekelleys.org.uk
>>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>>   
>>>       
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>   


From ranceh at gmail.com  Tue Nov 24 15:13:13 2009
From: ranceh at gmail.com (Rance Hall)
Date: Tue, 24 Nov 2009 09:13:13 -0600
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0BEDA0.1050304@jemms.net>
References: <4B06FE35.6010701@cyberspaceroad.com>
	<4B0AD436.5090605@cyberspaceroad.com> <4B0ADBC5.2000104@jemms.net>
	<4B0BC420.2000203@cyberspaceroad.com> <4B0BEDA0.1050304@jemms.net>
Message-ID: <845c0f80911240713p19e90305v6e84a483a8bcbc6f@mail.gmail.com>

On Tue, Nov 24, 2009 at 8:28 AM, Mark Beierl <mark at jemms.net> wrote:
> The TIME_WAIT is not an active socket, it's the remnant of a previous
> connection. ?I have no idea at all why mysqld has moved to 127.0.1.1.
> Is the bind address config line set to the host name and is the host
> name entry in /etc/hosts 127.0.1.1?
>
> Unfortunately, I know very little about mysql, so I can't point you in
> the right direction for configuration...
>
> Regards,
> Mark
>
> Adam Hardy wrote:
>> You're right. The result from netstat was:
>>
>>
>> tcp ? 0 ? 0 127.0.0.1:3306 ? 0.0.0.0:* ?LISTEN 2557/mysqld
>>
>> after changing the bindaddress config in the mysql config as per the docs to
>> free up networking, it then gives this result:
>>
>> adam at isengard:~$ sudo netstat -napt | grep 3306
>> tcp ? 0 ? 0 127.0.1.1:3306 ? ?0.0.0.0:* ? ? ? LISTEN ? ? ?16473/mysqld
>> tcp ? 0 ? 0 127.0.1.1:53067 ? 127.0.1.1:3306 ?TIME_WAIT ? -
>>
>> which looks weird. But then it's probably just because I don't have much
>> experience in this area. Why has it switched over to 127.0.1.1 ?
>>
>> Thanks
>> Adam
>>

The mysql configuration file is my.cnf
Its probably in /etc/ somewhere

By default mysql listens on the localhost interface only.

you can skip networking completely and just use pipes/sockets by using
the skip-networking directive in the my.cnf file.

Or you can add a bind-address directive to my.cnf.

bind-address 0.0.0.0  will listen on all interfaces
a series of bind-address x.x.x.x will specify a list of interfaces to
listen on  (don't forget to include localhost)

Rance


From adam.ant at cyberspaceroad.com  Tue Nov 24 17:47:36 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Tue, 24 Nov 2009 17:47:36 +0000
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0BEDA0.1050304@jemms.net>
References: <4B06FE35.6010701@cyberspaceroad.com>	<4B0AD436.5090605@cyberspaceroad.com>
	<4B0ADBC5.2000104@jemms.net> <4B0BC420.2000203@cyberspaceroad.com>
	<4B0BEDA0.1050304@jemms.net>
Message-ID: <4B0C1C38.8040904@cyberspaceroad.com>

I'm sorry, I must be quite annoying, giving stupid answers to the most basic 
networking questions. I checked in /etc/hosts:

127.0.0.1	localhost
127.0.1.1	isengard.localdomain	isengard

I didn't edit /etc/hosts myself so I'm not sure why 127.0.1.1 is in there. My lo 
interface according to ifconfig is 127.0.0.1

Before my previous email, I had also edited my mysql config via the my.cnf file 
to make the bind-address=isengard. I guess that explains the netstat output.

But I have no idea why 127.0.1.1 is there as well as 127.0.0.1 - what 
installation programs would have written that, other than the debian system 
install?

If it was just from the system installation, can I get rid of 127.0.1.1 and use 
all on one line:

127.0.0.1 localhost isengard.localdomain isengard

I also don't know whether to put some extra lines in my dnsmasq.conf and 
dhclient.conf, for instance in dnsmasq.conf the example

server=/localnet/192.168.0.1

looks useful, judging from the other stuff I just dealt with, although I'm not 
sure what purpose it serves.

The external NIC using dhclient.conf to get its IP address from the modem seems 
to be rewriting /etc/resolv.conf. Does dnsmasq deal with this on its own or is 
there a setting I should be using?

My dhclient.conf seems OK:

option rfc3442-classless-static-routes code 121 = array of unsigned integer 8;
request subnet-mask, broadcast-address, time-offset, routers,
	domain-name, domain-name-servers, domain-search, host-name,
	netbios-name-servers, netbios-scope, interface-mtu,
	rfc3442-classless-static-routes;


Regards
Adam



Mark Beierl on 24/11/09 14:28, wrote:
> The TIME_WAIT is not an active socket, it's the remnant of a previous 
> connection.  I have no idea at all why mysqld has moved to 127.0.1.1.  
> Is the bind address config line set to the host name and is the host 
> name entry in /etc/hosts 127.0.1.1?
> 
> Unfortunately, I know very little about mysql, so I can't point you in 
> the right direction for configuration...
> 
> Regards,
> Mark
> 
> Adam Hardy wrote:
>> You're right. The result from netstat was:
>>
>>
>> tcp   0   0 127.0.0.1:3306   0.0.0.0:*  LISTEN 2557/mysqld
>>
>> after changing the bindaddress config in the mysql config as per the 
>> docs to free up networking, it then gives this result:
>>
>> adam at isengard:~$ sudo netstat -napt | grep 3306
>> tcp   0   0 127.0.1.1:3306    0.0.0.0:*       LISTEN      16473/mysqld
>> tcp   0   0 127.0.1.1:53067   127.0.1.1:3306  TIME_WAIT   -
>>
>> which looks weird. But then it's probably just because I don't have 
>> much experience in this area. Why has it switched over to 127.0.1.1 ?
>>
>> Thanks
>> Adam
>>
>>
>> Mark Beierl on 23/11/09 19:00, wrote:
>>  
>>> Silly thought but - is mysql configured to listen to 127.0.0.1 only? 
>>> Something like
>>>
>>> sudo netstat -napt | grep 3306
>>>
>>> ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.
>>>
>>> Regards,
>>> Mark
>>>
>>> Adam Hardy wrote:
>>>    
>>>> Thought I had a simple problem but I don't really find anything 
>>>> relevant on the web and I'm not getting any responses to my 
>>>> questions here.
>>>>
>>>> Just a pointer in the right direction would be helpful - something 
>>>> to put me back on the scent?
>>>>
>>>> Thanks
>>>> Adam
>>>>
>>>> Adam Hardy on 20/11/09 20:38, wrote:
>>>>  
>>>>      
>>>>> I have a lan with a gateway machine running an ADSL modem and two 
>>>>> NICs with iptables and dnsmasq.
>>>>>
>>>>> It also runs mysql and tomcat but is currently just a simple 
>>>>> gateway, I'm not trying to configure any DMZ or fancier stuff like 
>>>>> that.
>>>>>
>>>>> My problem is that I can access mysql using 'localhost:3306' but I 
>>>>> can't access it on the same box when using the machine name e.g. 
>>>>> 'isengard:3306' and my guess is that I have mis-configured either 
>>>>> dnsmasq or iptables.
>>>>>
>>>>> I figure that my command "mysql --host=isengard" is probably being 
>>>>> resolved as external and then getting blocked by the firewall.
>>>>>
>>>>> My dnsmasq config file, based on the example config but with 
>>>>> comments removed, is:
>>>>>
>>>>> domain-needed
>>>>> bogus-priv
>>>>> filterwin2k
>>>>> local=/localdomain/
>>>>> domain=localdomain
>>>>> dhcp-range=192.168.0.3,192.168.0.254
>>>>> dhcp-option=option:router,192.168.0.2
>>>>> dhcp-option=option:mtu,1500
>>>>>
>>>>> and my resolv.conf file is:
>>>>>
>>>>> nameserver 194.74.65.68
>>>>>
>>>>> and I think this is getting continually rewritten by dhcp with the 
>>>>> nameserver info from the dhcp server on the modem which gives the 
>>>>> outside NIC its internet ip address.
>>>>>
>>>>> Does this make any sense? Or rather does anyone see where my 
>>>>> situation is foobarred?



From mark at jemms.net  Tue Nov 24 20:37:45 2009
From: mark at jemms.net (Mark Beierl)
Date: Tue, 24 Nov 2009 15:37:45 -0500
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0C1C38.8040904@cyberspaceroad.com>
References: <4B06FE35.6010701@cyberspaceroad.com>	<4B0AD436.5090605@cyberspaceroad.com>	<4B0ADBC5.2000104@jemms.net>
	<4B0BC420.2000203@cyberspaceroad.com>	<4B0BEDA0.1050304@jemms.net>
	<4B0C1C38.8040904@cyberspaceroad.com>
Message-ID: <4B0C4419.5010501@jemms.net>

Not a problem... just trying to help too :)

The 127.0.1.1 is a common thing these days, but I don't know why.  Yes 
you can merge them into one 127.0.0.1 line.

Rance Hall seemed to have hit the config entry on the head:

bind-address 0.0.0.0  

instead of isengard.  Due to isengard resolving to localhost/127.0.0.1 
(or 127.0.1.1 as the case may be), it is still the same as telling mysql 
to listen only to localhost, so that change made no real difference.  
You must tell mysql explicitly to listen to all addresses (0.0.0.0) with 
the bind-addresses listed above.

Regards,
Mark

Adam Hardy wrote:
> I'm sorry, I must be quite annoying, giving stupid answers to the most basic 
> networking questions. I checked in /etc/hosts:
>
> 127.0.0.1	localhost
> 127.0.1.1	isengard.localdomain	isengard
>
> I didn't edit /etc/hosts myself so I'm not sure why 127.0.1.1 is in there. My lo 
> interface according to ifconfig is 127.0.0.1
>
> Before my previous email, I had also edited my mysql config via the my.cnf file 
> to make the bind-address=isengard. I guess that explains the netstat output.
>
> But I have no idea why 127.0.1.1 is there as well as 127.0.0.1 - what 
> installation programs would have written that, other than the debian system 
> install?
>
> If it was just from the system installation, can I get rid of 127.0.1.1 and use 
> all on one line:
>
> 127.0.0.1 localhost isengard.localdomain isengard
>
> I also don't know whether to put some extra lines in my dnsmasq.conf and 
> dhclient.conf, for instance in dnsmasq.conf the example
>
> server=/localnet/192.168.0.1
>
> looks useful, judging from the other stuff I just dealt with, although I'm not 
> sure what purpose it serves.
>
> The external NIC using dhclient.conf to get its IP address from the modem seems 
> to be rewriting /etc/resolv.conf. Does dnsmasq deal with this on its own or is 
> there a setting I should be using?
>
> My dhclient.conf seems OK:
>
> option rfc3442-classless-static-routes code 121 = array of unsigned integer 8;
> request subnet-mask, broadcast-address, time-offset, routers,
> 	domain-name, domain-name-servers, domain-search, host-name,
> 	netbios-name-servers, netbios-scope, interface-mtu,
> 	rfc3442-classless-static-routes;
>
>
> Regards
> Adam
>
>
>
> Mark Beierl on 24/11/09 14:28, wrote:
>   
>> The TIME_WAIT is not an active socket, it's the remnant of a previous 
>> connection.  I have no idea at all why mysqld has moved to 127.0.1.1.  
>> Is the bind address config line set to the host name and is the host 
>> name entry in /etc/hosts 127.0.1.1?
>>
>> Unfortunately, I know very little about mysql, so I can't point you in 
>> the right direction for configuration...
>>
>> Regards,
>> Mark
>>
>> Adam Hardy wrote:
>>     
>>> You're right. The result from netstat was:
>>>
>>>
>>> tcp   0   0 127.0.0.1:3306   0.0.0.0:*  LISTEN 2557/mysqld
>>>
>>> after changing the bindaddress config in the mysql config as per the 
>>> docs to free up networking, it then gives this result:
>>>
>>> adam at isengard:~$ sudo netstat -napt | grep 3306
>>> tcp   0   0 127.0.1.1:3306    0.0.0.0:*       LISTEN      16473/mysqld
>>> tcp   0   0 127.0.1.1:53067   127.0.1.1:3306  TIME_WAIT   -
>>>
>>> which looks weird. But then it's probably just because I don't have 
>>> much experience in this area. Why has it switched over to 127.0.1.1 ?
>>>
>>> Thanks
>>> Adam
>>>
>>>
>>> Mark Beierl on 23/11/09 19:00, wrote:
>>>  
>>>       
>>>> Silly thought but - is mysql configured to listen to 127.0.0.1 only? 
>>>> Something like
>>>>
>>>> sudo netstat -napt | grep 3306
>>>>
>>>> ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.
>>>>
>>>> Regards,
>>>> Mark
>>>>
>>>> Adam Hardy wrote:
>>>>    
>>>>         
>>>>> Thought I had a simple problem but I don't really find anything 
>>>>> relevant on the web and I'm not getting any responses to my 
>>>>> questions here.
>>>>>
>>>>> Just a pointer in the right direction would be helpful - something 
>>>>> to put me back on the scent?
>>>>>
>>>>> Thanks
>>>>> Adam
>>>>>
>>>>> Adam Hardy on 20/11/09 20:38, wrote:
>>>>>  
>>>>>      
>>>>>           
>>>>>> I have a lan with a gateway machine running an ADSL modem and two 
>>>>>> NICs with iptables and dnsmasq.
>>>>>>
>>>>>> It also runs mysql and tomcat but is currently just a simple 
>>>>>> gateway, I'm not trying to configure any DMZ or fancier stuff like 
>>>>>> that.
>>>>>>
>>>>>> My problem is that I can access mysql using 'localhost:3306' but I 
>>>>>> can't access it on the same box when using the machine name e.g. 
>>>>>> 'isengard:3306' and my guess is that I have mis-configured either 
>>>>>> dnsmasq or iptables.
>>>>>>
>>>>>> I figure that my command "mysql --host=isengard" is probably being 
>>>>>> resolved as external and then getting blocked by the firewall.
>>>>>>
>>>>>> My dnsmasq config file, based on the example config but with 
>>>>>> comments removed, is:
>>>>>>
>>>>>> domain-needed
>>>>>> bogus-priv
>>>>>> filterwin2k
>>>>>> local=/localdomain/
>>>>>> domain=localdomain
>>>>>> dhcp-range=192.168.0.3,192.168.0.254
>>>>>> dhcp-option=option:router,192.168.0.2
>>>>>> dhcp-option=option:mtu,1500
>>>>>>
>>>>>> and my resolv.conf file is:
>>>>>>
>>>>>> nameserver 194.74.65.68
>>>>>>
>>>>>> and I think this is getting continually rewritten by dhcp with the 
>>>>>> nameserver info from the dhcp server on the modem which gives the 
>>>>>> outside NIC its internet ip address.
>>>>>>
>>>>>> Does this make any sense? Or rather does anyone see where my 
>>>>>> situation is foobarred?
>>>>>>             
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>   


From adam.ant at cyberspaceroad.com  Wed Nov 25 20:54:30 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Wed, 25 Nov 2009 20:54:30 +0000
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
In-Reply-To: <4B0C4419.5010501@jemms.net>
References: <4B06FE35.6010701@cyberspaceroad.com>	<4B0AD436.5090605@cyberspaceroad.com>	<4B0ADBC5.2000104@jemms.net>
	<4B0BC420.2000203@cyberspaceroad.com>	<4B0BEDA0.1050304@jemms.net>
	<4B0C1C38.8040904@cyberspaceroad.com> <4B0C4419.5010501@jemms.net>
Message-ID: <4B0D9986.8070807@cyberspaceroad.com>

OK I merged the hosts 127.0.1.1 names onto 127.0.0.1 with no ill effects so far, 
and with bind-address set to 0.0.0.0 I can see it listening to 0.0.0.0 also no 
ill effects. I just have to make sure now that I've got it covered from the 
outside world in iptables.

Thanks v. much for the help

Adam

Mark Beierl on 24/11/09 20:37, wrote:
> Not a problem... just trying to help too :)
> 
> The 127.0.1.1 is a common thing these days, but I don't know why.  Yes 
> you can merge them into one 127.0.0.1 line.
> 
> Rance Hall seemed to have hit the config entry on the head:
> 
> bind-address 0.0.0.0 
> instead of isengard.  Due to isengard resolving to localhost/127.0.0.1 
> (or 127.0.1.1 as the case may be), it is still the same as telling mysql 
> to listen only to localhost, so that change made no real difference.  
> You must tell mysql explicitly to listen to all addresses (0.0.0.0) with 
> the bind-addresses listed above.
> 
> Regards,
> Mark
> 
> Adam Hardy wrote:
>> I'm sorry, I must be quite annoying, giving stupid answers to the most 
>> basic networking questions. I checked in /etc/hosts:
>>
>> 127.0.0.1    localhost
>> 127.0.1.1    isengard.localdomain    isengard
>>
>> I didn't edit /etc/hosts myself so I'm not sure why 127.0.1.1 is in 
>> there. My lo interface according to ifconfig is 127.0.0.1
>>
>> Before my previous email, I had also edited my mysql config via the 
>> my.cnf file to make the bind-address=isengard. I guess that explains 
>> the netstat output.
>>
>> But I have no idea why 127.0.1.1 is there as well as 127.0.0.1 - what 
>> installation programs would have written that, other than the debian 
>> system install?
>>
>> If it was just from the system installation, can I get rid of 
>> 127.0.1.1 and use all on one line:
>>
>> 127.0.0.1 localhost isengard.localdomain isengard
>>
>> I also don't know whether to put some extra lines in my dnsmasq.conf 
>> and dhclient.conf, for instance in dnsmasq.conf the example
>>
>> server=/localnet/192.168.0.1
>>
>> looks useful, judging from the other stuff I just dealt with, although 
>> I'm not sure what purpose it serves.
>>
>> The external NIC using dhclient.conf to get its IP address from the 
>> modem seems to be rewriting /etc/resolv.conf. Does dnsmasq deal with 
>> this on its own or is there a setting I should be using?
>>
>> My dhclient.conf seems OK:
>>
>> option rfc3442-classless-static-routes code 121 = array of unsigned 
>> integer 8;
>> request subnet-mask, broadcast-address, time-offset, routers,
>>     domain-name, domain-name-servers, domain-search, host-name,
>>     netbios-name-servers, netbios-scope, interface-mtu,
>>     rfc3442-classless-static-routes;
>>
>>
>> Regards
>> Adam
>>
>>
>>
>> Mark Beierl on 24/11/09 14:28, wrote:
>>  
>>> The TIME_WAIT is not an active socket, it's the remnant of a previous 
>>> connection.  I have no idea at all why mysqld has moved to 
>>> 127.0.1.1.  Is the bind address config line set to the host name and 
>>> is the host name entry in /etc/hosts 127.0.1.1?
>>>
>>> Unfortunately, I know very little about mysql, so I can't point you 
>>> in the right direction for configuration...
>>>
>>> Regards,
>>> Mark
>>>
>>> Adam Hardy wrote:
>>>    
>>>> You're right. The result from netstat was:
>>>>
>>>>
>>>> tcp   0   0 127.0.0.1:3306   0.0.0.0:*  LISTEN 2557/mysqld
>>>>
>>>> after changing the bindaddress config in the mysql config as per the 
>>>> docs to free up networking, it then gives this result:
>>>>
>>>> adam at isengard:~$ sudo netstat -napt | grep 3306
>>>> tcp   0   0 127.0.1.1:3306    0.0.0.0:*       LISTEN      16473/mysqld
>>>> tcp   0   0 127.0.1.1:53067   127.0.1.1:3306  TIME_WAIT   -
>>>>
>>>> which looks weird. But then it's probably just because I don't have 
>>>> much experience in this area. Why has it switched over to 127.0.1.1 ?
>>>>
>>>> Thanks
>>>> Adam
>>>>
>>>>
>>>> Mark Beierl on 23/11/09 19:00, wrote:
>>>>  
>>>>      
>>>>> Silly thought but - is mysql configured to listen to 127.0.0.1 
>>>>> only? Something like
>>>>>
>>>>> sudo netstat -napt | grep 3306
>>>>>
>>>>> ought to show if mysql is listening on 127.0.0.1:3306 or 0.0.0.0:3306.
>>>>>
>>>>> Regards,
>>>>> Mark
>>>>>
>>>>> Adam Hardy wrote:
>>>>>           
>>>>>> Thought I had a simple problem but I don't really find anything 
>>>>>> relevant on the web and I'm not getting any responses to my 
>>>>>> questions here.
>>>>>>
>>>>>> Just a pointer in the right direction would be helpful - something 
>>>>>> to put me back on the scent?
>>>>>>
>>>>>> Thanks
>>>>>> Adam
>>>>>>
>>>>>> Adam Hardy on 20/11/09 20:38, wrote:
>>>>>>  
>>>>>>               
>>>>>>> I have a lan with a gateway machine running an ADSL modem and two 
>>>>>>> NICs with iptables and dnsmasq.
>>>>>>>
>>>>>>> It also runs mysql and tomcat but is currently just a simple 
>>>>>>> gateway, I'm not trying to configure any DMZ or fancier stuff 
>>>>>>> like that.
>>>>>>>
>>>>>>> My problem is that I can access mysql using 'localhost:3306' but 
>>>>>>> I can't access it on the same box when using the machine name 
>>>>>>> e.g. 'isengard:3306' and my guess is that I have mis-configured 
>>>>>>> either dnsmasq or iptables.
>>>>>>>
>>>>>>> I figure that my command "mysql --host=isengard" is probably 
>>>>>>> being resolved as external and then getting blocked by the firewall.
>>>>>>>
>>>>>>> My dnsmasq config file, based on the example config but with 
>>>>>>> comments removed, is:
>>>>>>>
>>>>>>> domain-needed
>>>>>>> bogus-priv
>>>>>>> filterwin2k
>>>>>>> local=/localdomain/
>>>>>>> domain=localdomain
>>>>>>> dhcp-range=192.168.0.3,192.168.0.254
>>>>>>> dhcp-option=option:router,192.168.0.2
>>>>>>> dhcp-option=option:mtu,1500
>>>>>>>
>>>>>>> and my resolv.conf file is:
>>>>>>>
>>>>>>> nameserver 194.74.65.68
>>>>>>>
>>>>>>> and I think this is getting continually rewritten by dhcp with 
>>>>>>> the nameserver info from the dhcp server on the modem which gives 
>>>>>>> the outside NIC its internet ip address.
>>>>>>>
>>>>>>> Does this make any sense? Or rather does anyone see where my 
>>>>>>> situation is foobarred?



From adam.ant at cyberspaceroad.com  Thu Nov 26 19:00:12 2009
From: adam.ant at cyberspaceroad.com (Adam Hardy)
Date: Thu, 26 Nov 2009 19:00:12 +0000 (UTC)
Subject: [Dnsmasq-discuss] dnsmasq / dns server / iptables config glitch?
References: <4B06FE35.6010701@cyberspaceroad.com>
	<4B0AD436.5090605@cyberspaceroad.com>
	<007e01ca6c83$61368900$23a39b00$@net>
Message-ID: <loom.20091126T195504-222@post.gmane.org>

Brad Morgan <b-morgan at ...> writes:
> I have a very similar configuration. I think you are close but you may 
> to tweak your DHCP client. 
[snip]
> Take a look at man dhclient.conf to see what might make sense for your
> configuration. I think "prepend domain-name-servers 127.0.0.1" may be 
> you need. 
> 
> My /etc/dhclient-eth0.conf:
> 
> supersede domain-name "morgan.local"; 
> supersede domain-name-servers 127.0.0.1,208.67.222.222,208.67.220.220;
> #OpenDNS
> send host-name "bricknix";  # temporary RHL ifup addition

Actually I didn't realise how important those supersede commands were, but
without them, the clients on my network won't resolve the actual dnsmasq 
machine. 

The windows machine kept resolving the hostname isengard (name of the dnsm
machine) to 127.0.0.1! Pretty confusing, and nslookup on the windows machi
said something about non-existent domains and PTR records. 

Anyway, all ship shape now. 
Ta
Adam



From laci at monda.hu  Sat Nov 28 17:44:00 2009
From: laci at monda.hu (=?UTF-8?B?TMOhc3psw7MgTW9uZGE=?=)
Date: Sat, 28 Nov 2009 18:44:00 +0100
Subject: [Dnsmasq-discuss] Using the same IP for the wired and wireless
	interface of a laptop
Message-ID: <4d598e970911280944n43c663c5qd187d24711afaebd@mail.gmail.com>

Hi List,

I'm trying to bind two MACs which are the wired and the wireless
interface of a laptop to one IP by specifying two related lines in
/etc/ethers and one related line in /etc/hosts but it seems only the
first MAC works.  I guess dnsmasq overrides the first with the second.

It'd be nice to use the same IP for the laptop regardless whether the
wired or the wireless interface used, altough I'm not sure that this
is a really good idea because when both of the interfaces are active
this can lead to troubles.

What do you think?

-- 
Laci  <http://monda.hu>


From v_cadet at yahoo.fr  Sat Nov 28 22:27:20 2009
From: v_cadet at yahoo.fr (Vincent Cadet)
Date: Sat, 28 Nov 2009 22:27:20 +0000 (GMT)
Subject: [Dnsmasq-discuss] Re : Using the same IP for the wired and wireless
	interface of a laptop
In-Reply-To: <4d598e970911280944n43c663c5qd187d24711afaebd@mail.gmail.com>
Message-ID: <822911.22156.qm@web28202.mail.ukl.yahoo.com>

Hi  L?szl?.

> I'm trying to bind two MACs which are the wired and the
> wireless
> interface of a laptop to one IP by specifying two related
> lines in
> /etc/ethers and one related line in /etc/hosts but it seems
> only the
> first MAC works.  I guess dnsmasq overrides the first
> with the second.
> 
> It'd be nice to use the same IP for the laptop regardless
> whether the
> wired or the wireless interface used, altough I'm not sure
> that this
> is a really good idea because when both of the interfaces
> are active
> this can lead to troubles.
> 
> What do you think?

You can use bonding for that. I've successfully used bonding with my wired and wireless interfaces on my laptop. The result is one virtual interface (hece one IP address) that combines both physical interfaces w/ possible failover or load balancing.

Vince C.


      


From laci at monda.hu  Mon Nov 30 01:53:26 2009
From: laci at monda.hu (=?UTF-8?B?TMOhc3psw7MgTW9uZGE=?=)
Date: Mon, 30 Nov 2009 02:53:26 +0100
Subject: [Dnsmasq-discuss] Re : Using the same IP for the wired and
	wireless interface of a laptop
In-Reply-To: <822911.22156.qm@web28202.mail.ukl.yahoo.com>
References: <4d598e970911280944n43c663c5qd187d24711afaebd@mail.gmail.com>
	<822911.22156.qm@web28202.mail.ukl.yahoo.com>
Message-ID: <4d598e970911291753w1e41f4b8sbcf3f66407ce7c3c@mail.gmail.com>

Hi Vincent,

On Sat, Nov 28, 2009 at 11:27 PM, Vincent Cadet <v_cadet at yahoo.fr> wrote:
> Hi ?L?szl?.
>
>> I'm trying to bind two MACs which are the wired and the
>> wireless
>> interface of a laptop to one IP by specifying two related
>> lines in
>> /etc/ethers and one related line in /etc/hosts but it seems
>> only the
>> first MAC works. ?I guess dnsmasq overrides the first
>> with the second.
>>
>> It'd be nice to use the same IP for the laptop regardless
>> whether the
>> wired or the wireless interface used, altough I'm not sure
>> that this
>> is a really good idea because when both of the interfaces
>> are active
>> this can lead to troubles.
>>
>> What do you think?
>
> You can use bonding for that. I've successfully used bonding with my wired and wireless interfaces on my laptop. The result is one virtual interface (hece one IP address) that combines both physical interfaces w/ possible failover or load balancing.

Great idea!  Bonding is much cleaner than my solution.

Thanks!

-- 
Laci  <http://monda.hu>


From simon at thekelleys.org.uk  Mon Nov 30 17:23:52 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 30 Nov 2009 17:23:52 +0000
Subject: [Dnsmasq-discuss] Using the same IP for the wired and wireless
 interface of a laptop
In-Reply-To: <4d598e970911280944n43c663c5qd187d24711afaebd@mail.gmail.com>
References: <4d598e970911280944n43c663c5qd187d24711afaebd@mail.gmail.com>
Message-ID: <4B13FFA8.6060806@thekelleys.org.uk>

L?szl? Monda wrote:
> Hi List,
> 
> I'm trying to bind two MACs which are the wired and the wireless
> interface of a laptop to one IP by specifying two related lines in
> /etc/ethers and one related line in /etc/hosts but it seems only the
> first MAC works.  I guess dnsmasq overrides the first with the second.
> 
> It'd be nice to use the same IP for the laptop regardless whether the
> wired or the wireless interface used, altough I'm not sure that this
> is a really good idea because when both of the interfaces are active
> this can lead to troubles.
> 
> What do you think?
> 

You can make dnsmasq share the same IP address between two interfaces by 
putting both MAc addresses on one line in /etc/dnsmasq.conf

dhcp-host=<MAC 1>,<MAC 2>,<IP-address>

This will only work if you're careful to make sure that only one 
interface at a time is up.

Cheers,

Simon.




From marc at marcsallent.com  Wed Dec  2 09:53:03 2009
From: marc at marcsallent.com (Marc Sallent)
Date: Wed, 2 Dec 2009 10:53:03 +0100
Subject: [Dnsmasq-discuss] Resolve all domains to the same IP
Message-ID: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>

Hi!

For an interactive installation, I need everybody connecting to my wi- 
fi network to be shown the same page, hosted in my computer.
I have a router with Dnsmasq installed, connected to the computer with  
the server, IP 1.2.3.4, so everybody looking for, ie. google.com,  
yahoo.com should see only the contents in the server with IP 1.2.3.4

Is that possible?

Thanks!

marc


From simon at thekelleys.org.uk  Wed Dec  2 09:57:04 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 02 Dec 2009 09:57:04 +0000
Subject: [Dnsmasq-discuss] Resolve all domains to the same IP
In-Reply-To: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
References: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
Message-ID: <4B1639F0.8080605@thekelleys.org.uk>

Marc Sallent wrote:
> Hi!
> 
> For an interactive installation, I need everybody connecting to my wi- 
> fi network to be shown the same page, hosted in my computer.
> I have a router with Dnsmasq installed, connected to the computer with  
> the server, IP 1.2.3.4, so everybody looking for, ie. google.com,  
> yahoo.com should see only the contents in the server with IP 1.2.3.4
> 
> Is that possible?
> 
> Thanks!
> 
> marc
> 
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
> 


--address =/#/1.2.3.4

Check the man page for more details.


HTH


Simon.





From laci at monda.hu  Wed Dec  2 10:21:27 2009
From: laci at monda.hu (=?UTF-8?B?TMOhc3psw7MgTW9uZGE=?=)
Date: Wed, 2 Dec 2009 11:21:27 +0100
Subject: [Dnsmasq-discuss] Resolve all domains to the same IP
In-Reply-To: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
References: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
Message-ID: <4d598e970912020221i45c4da8ftd189fa9bd15ad270@mail.gmail.com>

Hi,

First of all, using 1.2.3.4 as a private IP address is a very bad
practice because it's not part of the private IP address space -
http://en.wikipedia.org/wiki/Private_network

Second, what are you really trying to do?  If you want to set up a
captive portal on OpenWrt then you have some options:

root at sunshine:~# opkg list|grep -i captive
chillispot - 1.1.0-1 - ChilliSpot is an open source captive portal or
wireless LAN
pepperspot - 0.1-1 - PepperSpot is a captive portal or wireless LAN
access point controller
wifidog - 1.1.5-2 - The Wifidog project is a complete and embeddable captive

Please CC me in your answer because I won't be on the list.

On Wed, Dec 2, 2009 at 10:53 AM, Marc Sallent <marc at marcsallent.com> wrote:
> Hi!
>
> For an interactive installation, I need everybody connecting to my wi-
> fi network to be shown the same page, hosted in my computer.
> I have a router with Dnsmasq installed, connected to the computer with
> the server, IP 1.2.3.4, so everybody looking for, ie. google.com,
> yahoo.com should see only the contents in the server with IP 1.2.3.4
>
> Is that possible?
>
> Thanks!
>
> marc
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>



-- 
Laci  <http://monda.hu>


From laci at monda.hu  Wed Dec  2 11:23:46 2009
From: laci at monda.hu (=?UTF-8?B?TMOhc3psw7MgTW9uZGE=?=)
Date: Wed, 2 Dec 2009 12:23:46 +0100
Subject: [Dnsmasq-discuss] Resolve all domains to the same IP
In-Reply-To: <4B1639F0.8080605@thekelleys.org.uk>
References: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
	<4B1639F0.8080605@thekelleys.org.uk>
Message-ID: <4d598e970912020323y314edfa7r6ec4e5fbb1883274@mail.gmail.com>

On Wed, Dec 2, 2009 at 10:57 AM, Simon Kelley <simon at thekelleys.org.uk> wrote:
> Marc Sallent wrote:
>> Hi!
>>
>> For an interactive installation, I need everybody connecting to my wi-
>> fi network to be shown the same page, hosted in my computer.
>> I have a router with Dnsmasq installed, connected to the computer with
>> the server, IP 1.2.3.4, so everybody looking for, ie. google.com,
>> yahoo.com should see only the contents in the server with IP 1.2.3.4
>>
>> Is that possible?
>>
>> Thanks!
>>
>> marc
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>
>
> --address =/#/1.2.3.4
>
> Check the man page for more details.
>
>
> HTH
>
>
> Simon.
>

Now I see what it's all about.  Thanks for clarifying.

>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>



-- 
Laci  <http://monda.hu>


From v_cadet at yahoo.fr  Wed Dec  2 15:14:06 2009
From: v_cadet at yahoo.fr (Vincent Cadet)
Date: Wed, 2 Dec 2009 15:14:06 +0000 (GMT)
Subject: [Dnsmasq-discuss] Re :  Resolve all domains to the same IP
In-Reply-To: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
Message-ID: <242697.34737.qm@web28207.mail.ukl.yahoo.com>

Hello Marc

> For an interactive installation, I need everybody
> connecting to my wi- 
> fi network to be shown the same page, hosted in my
> computer.
> I have a router with Dnsmasq installed, connected to the
> computer with  
> the server, IP 1.2.3.4, so everybody looking for, ie.
> google.com,  
> yahoo.com should see only the contents in the server with
> IP 1.2.3.4
> 
> Is that possible?

As L?szl? Monda wrote, what you are looking for is a captive portal, such as Chillispot (et al.) Such applications are really dedicated to providing a web page where visitors must enter a login and a password to continue surfing.

You can use dnsmasq to send visitors to the same IP address. Unlike a captive portal it can't be dynamic as you'd have to stop the redirection as soon as users are identified, which can't be done easily with dnsmasq and most probably not on a per-user basis. Moreover you'd permanently have every network packet sent to the same IP address, not just web traffic.

Hope this helps,
Vince C.


      


From elaganowski at hotmail.com  Wed Dec  2 16:45:57 2009
From: elaganowski at hotmail.com (Eric Laganowski)
Date: Wed, 2 Dec 2009 11:45:57 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
Message-ID: <SNT130-w394FD18999D0797477ACB7A6950@phx.gbl>


Hi! 

I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow. 
Is it possible to accomplish this with dnsmasq? 

-Eric  		 	   		  
_________________________________________________________________
Chat with Messenger straight from your Hotmail inbox.
http://www.microsoft.com/windows/windowslive/hotmail_bl1/hotmail_bl1.aspx?ocid=PID23879::T:WLMTAGL:ON:WL:en-ww:WM_IMHM_4:092009
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091202/b234dc7f/attachment.htm 

From santiago at zarate.net.ve  Wed Dec  2 17:22:35 2009
From: santiago at zarate.net.ve (Santiago Zarate)
Date: Wed, 2 Dec 2009 12:52:35 -0430
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <SNT130-w394FD18999D0797477ACB7A6950@phx.gbl>
References: <SNT130-w394FD18999D0797477ACB7A6950@phx.gbl>
Message-ID: <200912021252.35536.santiago@zarate.net.ve>

i guess you can use a cname record...
address=/wpad.mydomain.net/10.40.60.90
cname=wpad,wpad.mydomain.net

tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work... 

-- 
Santiago Zarate
santiago at zarate.net.ve
(+58) 4129864175
(+58) 4241073905


From steve at thestever.net  Wed Dec  2 17:32:07 2009
From: steve at thestever.net (Steve Quezadas)
Date: Wed, 02 Dec 2009 09:32:07 -0800
Subject: [Dnsmasq-discuss] windows' hostname
Message-ID: <4B16A497.8000205@thestever.net>

Simple question:

Can dnsmasq set the hostname for a windows station? I ghosted an image 
across 5 computers and now they all have the same hostname. I am hoping 
that dhcp allows you to set the hostname of each machine by mac address 
so I won't get the "a duplicate name exists on the network" error.

- Steve



From b-morgan at concentric.net  Wed Dec  2 17:50:46 2009
From: b-morgan at concentric.net (Brad Morgan)
Date: Wed, 2 Dec 2009 10:50:46 -0700
Subject: [Dnsmasq-discuss] windows' hostname
In-Reply-To: <4B16A497.8000205@thestever.net>
References: <4B16A497.8000205@thestever.net>
Message-ID: <008501ca7377$fb41b3d0$f1c51b70$@net>

> Can dnsmasq set the hostname for a windows station? I ghosted an image 
> across 5 computers and now they all have the same hostname. I am hoping 
> that dhcp allows you to set the hostname of each machine by mac address 
> so I won't get the "a duplicate name exists on the network" error.

You are going to have more problems than just the hostname with your Windows
systems. They also have duplicate system identifications (SIDs). You will
need to run a program like NewSID
(http://technet.microsoft.com/en-us/sysinternals/bb897418.aspx) on each
system. You can change the hostname as well at this time.

Regards,

Brad




From nikolay.kasatkin at gmail.com  Wed Dec  2 18:09:14 2009
From: nikolay.kasatkin at gmail.com (Nikolay Kasatkin)
Date: Wed, 2 Dec 2009 21:09:14 +0300
Subject: [Dnsmasq-discuss] windows' hostname
In-Reply-To: <4B16A497.8000205@thestever.net>
References: <4B16A497.8000205@thestever.net>
Message-ID: <78710f9a0912021009v650a59dfge0281f83a790e858@mail.gmail.com>

Give a try to Clonezilla and DRBL (clonezilla.org and
http://drbl.sourceforge.net). They can change SID and windows hostname after
cloning too

On Wed, Dec 2, 2009 at 8:32 PM, Steve Quezadas <steve at thestever.net> wrote:

> Simple question:
>
> Can dnsmasq set the hostname for a windows station? I ghosted an image
> across 5 computers and now they all have the same hostname. I am hoping
> that dhcp allows you to set the hostname of each machine by mac address
> so I won't get the "a duplicate name exists on the network" error.
>
> - Steve
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>



-- 
Kasatkin Nikolay
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091202/39808ae9/attachment.htm 

From Hullen at t-online.de  Wed Dec  2 18:28:00 2009
From: Hullen at t-online.de (Helmut Hullen)
Date: 02 Dec 2009 19:28:00 +0100
Subject: [Dnsmasq-discuss] windows' hostname
In-Reply-To: <008501ca7377$fb41b3d0$f1c51b70$@net>
Message-ID: <BE73kTQzSsB@helmut.hullen.de>

Hallo, Brad,

Du meintest am 02.12.09:

>> Can dnsmasq set the hostname for a windows station? I ghosted an
>> image across 5 computers and now they all have the same hostname.

[...]

> You are going to have more problems than just the hostname with your
> Windows systems. They also have duplicate system identifications
> (SIDs). You will need to run a program like NewSID
> (http://technet.microsoft.com/en-us/sysinternals/bb897418.aspx) on
> each system. You can change the hostname as well at this time.

Take a look at

<http://blogs.technet.com/markrussinovich/archive/2009/11/03/3291024.aspx>

especially "SID Duplication"

(ok - it's not related in any way to dnsmasq)


> Regards,

> Brad



> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss


Viele Gruesse!
Helmut


From joshua at eeinternet.com  Wed Dec  2 19:07:55 2009
From: joshua at eeinternet.com (Joshua J. Kugler)
Date: Wed, 02 Dec 2009 10:07:55 -0900
Subject: [Dnsmasq-discuss] windows' hostname
In-Reply-To: <008501ca7377$fb41b3d0$f1c51b70$@net>
References: <4B16A497.8000205@thestever.net>
	<008501ca7377$fb41b3d0$f1c51b70$@net>
Message-ID: <200912021007.56160.joshua@eeinternet.com>

On Wednesday 02 December 2009, Brad Morgan said something like:
> > Can dnsmasq set the hostname for a windows station? I ghosted an
> > image across 5 computers and now they all have the same hostname. I
> > am hoping that dhcp allows you to set the hostname of each machine
> > by mac address so I won't get the "a duplicate name exists on the
> > network" error.
>
> You are going to have more problems than just the hostname with your
> Windows systems. They also have duplicate system identifications
> (SIDs). You will need to run a program like NewSID
> (http://technet.microsoft.com/en-us/sysinternals/bb897418.aspx) on
> each system. You can change the hostname as well at this time.

The SID thing appears to be no longer true. See 
http://blogs.technet.com/markrussinovich/archive/2009/11/03/3291024.aspx

j

-- 
Joshua Kugler
Part-Time System Admin/Programmer
http://www.eeinternet.com
PGP Key: http://pgp.mit.edu/ ?ID 0x14EA086E



From richardvoigt at gmail.com  Thu Dec  3 02:50:07 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Wed, 2 Dec 2009 20:50:07 -0600
Subject: [Dnsmasq-discuss] Resolve all domains to the same IP
In-Reply-To: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
References: <05386E92-A9BD-4AF0-8123-6A5482D3045B@marcsallent.com>
Message-ID: <2e59e6970912021850o7d01e98cq1f3aecade953eaf@mail.gmail.com>

On Wed, Dec 2, 2009 at 3:53 AM, Marc Sallent <marc at marcsallent.com> wrote:
> Hi!
>
> For an interactive installation, I need everybody connecting to my wi-
> fi network to be shown the same page, hosted in my computer.
> I have a router with Dnsmasq installed, connected to the computer with
> the server, IP 1.2.3.4, so everybody looking for, ie. google.com,
> yahoo.com should see only the contents in the server with IP 1.2.3.4
>
> Is that possible?

No, it is not possible.  You cannot force people to view a web page.
You cannot open a web browser remotely.  Even if they open the browser
voluntarily, DNS will not help you since it only is involved when the
client uses a name.  If the client connects by IP address (common
since TCP/IP stacks cache recently used lookups) then dnsmasq never
enters the picture.

Use masquerade as suggested by Laszlo, and realize that people will
still not see your page until they choose to open their browser.

>
> Thanks!
>
> marc
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eric at laganowski.net  Thu Dec  3 22:48:52 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Thu, 03 Dec 2009 17:48:52 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
Message-ID: <4B184054.80408@laganowski.net>

I guess I am not communicating this well.
The desired dnsmasq behavior would be to reply, say, with 192.168.1.1 to any request starting with "wpad.", not just local domain, so if, say my domain is domain.local, dnsmasq responds with 192.168.1.1 to both wpad.domain.local and wpad.google.com

-Eric

Date: Wed, 2 Dec 2009 12:52:35 -0430
From: Santiago Zarate <santiago at zarate.net.ve>
Subject: Re: [Dnsmasq-discuss] DNS pattern response
To: dnsmasq-discuss at lists.thekelleys.org.uk
Message-ID: <200912021252.35536.santiago at zarate.net.ve>
Content-Type: Text/Plain;  charset="us-ascii"

i guess you can use a cname record...
address=/wpad.mydomain.net/10.40.60.90
cname=wpad,wpad.mydomain.net

tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work... 


-- Santiago Zarate santiago at zarate.net.ve (+58) 4129864175 (+58) 4241073905


Date: Wed, 2 Dec 2009 11:45:57 -0500
From: Eric Laganowski <elaganowski at hotmail.com>
Subject: [Dnsmasq-discuss] DNS pattern response
To: <dnsmasq-discuss at lists.thekelleys.org.uk>
Message-ID: <SNT130-w394FD18999D0797477ACB7A6950 at phx.gbl>
Content-Type: text/plain; charset="iso-8859-1"


Hi! 

I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow. 
Is it possible to accomplish this with dnsmasq? 

-Eric 



From tmetro+dnsmasq at gmail.com  Thu Dec  3 23:41:19 2009
From: tmetro+dnsmasq at gmail.com (Tom Metro)
Date: Thu, 03 Dec 2009 18:41:19 -0500
Subject: [Dnsmasq-discuss] query rewriting
Message-ID: <4B184C9F.1040705@gmail.com>

Is there a dnsmasq directive similar to 'alias' that instead of 
rewriting IP addresses in the query reply, rewrites the domain in the 
query? So a query for foo.domain always results in a query for bar.domain?

This is related to the idea of a CNAME, but the dnsmasq 'cname' 
directive is restricted to bar.domain being local. I'm looking for 
something that will work for domains that aren't locally controlled. The 
'address' directive is also related, but the target needs to be a 
domain, or at least a cached lookup, and not a static IP.

  -Tom

-- 
Tom Metro
Venture Logic, Newton, MA, USA
"Enterprise solutions through open source."
Professional Profile: http://tmetro.venturelogic.com/


From richardvoigt at gmail.com  Fri Dec  4 00:17:06 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Thu, 3 Dec 2009 18:17:06 -0600
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B184054.80408@laganowski.net>
References: <4B184054.80408@laganowski.net>
Message-ID: <2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>

I think you're going about this the wrong way.  Use the DHCP option
"auto-proxy-config" to control the URL browsers use for auto-proxy.
Spoofing addresses in other domains doesn't solve any problems, it
only creates more (and is borderline illegal in many areas).

On Thu, Dec 3, 2009 at 4:48 PM, Eric Laganowski <eric at laganowski.net> wrote:
> I guess I am not communicating this well.
> The desired dnsmasq behavior would be to reply, say, with 192.168.1.1 to any request starting with "wpad.", not just local domain, so if, say my domain is domain.local, dnsmasq responds with 192.168.1.1 to both wpad.domain.local and wpad.google.com
>
> -Eric
>
> Date: Wed, 2 Dec 2009 12:52:35 -0430
> From: Santiago Zarate <santiago at zarate.net.ve>
> Subject: Re: [Dnsmasq-discuss] DNS pattern response
> To: dnsmasq-discuss at lists.thekelleys.org.uk
> Message-ID: <200912021252.35536.santiago at zarate.net.ve>
> Content-Type: Text/Plain; ?charset="us-ascii"
>
> i guess you can use a cname record...
> address=/wpad.mydomain.net/10.40.60.90
> cname=wpad,wpad.mydomain.net
>
> tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work...
>
>
> -- Santiago Zarate santiago at zarate.net.ve (+58) 4129864175 (+58) 4241073905
>
>
> Date: Wed, 2 Dec 2009 11:45:57 -0500
> From: Eric Laganowski <elaganowski at hotmail.com>
> Subject: [Dnsmasq-discuss] DNS pattern response
> To: <dnsmasq-discuss at lists.thekelleys.org.uk>
> Message-ID: <SNT130-w394FD18999D0797477ACB7A6950 at phx.gbl>
> Content-Type: text/plain; charset="iso-8859-1"
>
>
> Hi!
>
> I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow.
> Is it possible to accomplish this with dnsmasq?
>
> -Eric
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From richardvoigt at gmail.com  Fri Dec  4 00:18:35 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Thu, 3 Dec 2009 18:18:35 -0600
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <4B184C9F.1040705@gmail.com>
References: <4B184C9F.1040705@gmail.com>
Message-ID: <2e59e6970912031618j552db0f5y83a31aeee530e13@mail.gmail.com>

Spoofing addresses in other people's domains doesn't solve any
problems, it only creates more (and is borderline illegal in many
areas).

What are you trying to accomplish?

On Thu, Dec 3, 2009 at 5:41 PM, Tom Metro <tmetro+dnsmasq at gmail.com> wrote:
> Is there a dnsmasq directive similar to 'alias' that instead of
> rewriting IP addresses in the query reply, rewrites the domain in the
> query? So a query for foo.domain always results in a query for bar.domain?
>
> This is related to the idea of a CNAME, but the dnsmasq 'cname'
> directive is restricted to bar.domain being local. I'm looking for
> something that will work for domains that aren't locally controlled. The
> 'address' directive is also related, but the target needs to be a
> domain, or at least a cached lookup, and not a static IP.
>
> ?-Tom
>
> --
> Tom Metro
> Venture Logic, Newton, MA, USA
> "Enterprise solutions through open source."
> Professional Profile: http://tmetro.venturelogic.com/
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eric at laganowski.net  Fri Dec  4 03:45:26 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Thu, 03 Dec 2009 22:45:26 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
Message-ID: <4B1885D6.1050105@laganowski.net>

Well, while legal advice is always appreciated that was not what I was 
asking for.
I was asking about a specific feature of dnsmasq and I am still at a 
loss whether it is possible to accomplish what I was looking for or not.

On a side note, DHCP option 252 is not supported by Firefox, that is way 
I am forced to explore other options.

Thanks,
Eric

richardvoigt at gmail.com wrote:
> I think you're going about this the wrong way.  Use the DHCP option
> "auto-proxy-config" to control the URL browsers use for auto-proxy.
> Spoofing addresses in other domains doesn't solve any problems, it
> only creates more (and is borderline illegal in many areas).
>
> On Thu, Dec 3, 2009 at 4:48 PM, Eric Laganowski <eric at laganowski.net> wrote:
>   
>> I guess I am not communicating this well.
>> The desired dnsmasq behavior would be to reply, say, with 192.168.1.1 to any request starting with "wpad.", not just local domain, so if, say my domain is domain.local, dnsmasq responds with 192.168.1.1 to both wpad.domain.local and wpad.google.com
>>
>> -Eric
>>
>> Date: Wed, 2 Dec 2009 12:52:35 -0430
>> From: Santiago Zarate <santiago at zarate.net.ve>
>> Subject: Re: [Dnsmasq-discuss] DNS pattern response
>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>> Message-ID: <200912021252.35536.santiago at zarate.net.ve>
>> Content-Type: Text/Plain;  charset="us-ascii"
>>
>> i guess you can use a cname record...
>> address=/wpad.mydomain.net/10.40.60.90
>> cname=wpad,wpad.mydomain.net
>>
>> tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work...
>>
>>
>> -- Santiago Zarate santiago at zarate.net.ve (+58) 4129864175 (+58) 4241073905
>>
>>
>> Date: Wed, 2 Dec 2009 11:45:57 -0500
>> From: Eric Laganowski <elaganowski at hotmail.com>
>> Subject: [Dnsmasq-discuss] DNS pattern response
>> To: <dnsmasq-discuss at lists.thekelleys.org.uk>
>> Message-ID: <SNT130-w394FD18999D0797477ACB7A6950 at phx.gbl>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>>
>> Hi!
>>
>> I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow.
>> Is it possible to accomplish this with dnsmasq?
>>
>> -Eric
>>
>>
>> _______________________________________________
>> Dnsmasq-discuss mailing list
>> Dnsmasq-discuss at lists.thekelleys.org.uk
>> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>>
>>     



From perette at barella.org  Fri Dec  4 04:06:32 2009
From: perette at barella.org (Perette Barella)
Date: Thu, 3 Dec 2009 23:06:32 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B1885D6.1050105@laganowski.net>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
Message-ID: <A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>

I think there's a misunderstanding on how the WPAD DNS version operates. The "wpad.domain.localnet" is used by the browser at startup to locate the proxy configuration file which applies to all domains.  You don't need a separate wpad.google.com and wpad.amazon.com for every domain users are trying to connect to.

If for some reason your local hosts are configured with different domain names (and therefore looking up wpad.google.com or wpad.amazon.com), I think we need more explanation on just what strangeness you've got going on.

Perette


You can do the same thing by 
On 2009?12?03?, at 22:45, Eric Laganowski wrote:

> Well, while legal advice is always appreciated that was not what I was 
> asking for.
> I was asking about a specific feature of dnsmasq and I am still at a 
> loss whether it is possible to accomplish what I was looking for or not.
> 
> On a side note, DHCP option 252 is not supported by Firefox, that is way 
> I am forced to explore other options.
> 
> Thanks,
> Eric
> 
> richardvoigt at gmail.com wrote:
>> I think you're going about this the wrong way.  Use the DHCP option
>> "auto-proxy-config" to control the URL browsers use for auto-proxy.
>> Spoofing addresses in other domains doesn't solve any problems, it
>> only creates more (and is borderline illegal in many areas).
>> 
>> On Thu, Dec 3, 2009 at 4:48 PM, Eric Laganowski <eric at laganowski.net> wrote:
>> 
>>> I guess I am not communicating this well.
>>> The desired dnsmasq behavior would be to reply, say, with 192.168.1.1 to any request starting with "wpad.", not just local domain, so if, say my domain is domain.local, dnsmasq responds with 192.168.1.1 to both wpad.domain.local and wpad.google.com
>>> 
>>> -Eric
>>> 
>>> Date: Wed, 2 Dec 2009 12:52:35 -0430
>>> From: Santiago Zarate <santiago at zarate.net.ve>
>>> Subject: Re: [Dnsmasq-discuss] DNS pattern response
>>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>>> Message-ID: <200912021252.35536.santiago at zarate.net.ve>
>>> Content-Type: Text/Plain;  charset="us-ascii"
>>> 
>>> i guess you can use a cname record...
>>> address=/wpad.mydomain.net/10.40.60.90
>>> cname=wpad,wpad.mydomain.net
>>> 
>>> tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work...
>>> 
>>> 
>>> -- Santiago Zarate santiago at zarate.net.ve (+58) 4129864175 (+58) 4241073905
>>> 
>>> 
>>> Date: Wed, 2 Dec 2009 11:45:57 -0500
>>> From: Eric Laganowski <elaganowski at hotmail.com>
>>> Subject: [Dnsmasq-discuss] DNS pattern response
>>> To: <dnsmasq-discuss at lists.thekelleys.org.uk>
>>> Message-ID: <SNT130-w394FD18999D0797477ACB7A6950 at phx.gbl>
>>> Content-Type: text/plain; charset="iso-8859-1"
>>> 
>>> 
>>> Hi!
>>> 
>>> I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow.
>>> Is it possible to accomplish this with dnsmasq?
>>> 
>>> -Eric



From simon at thekelleys.org.uk  Fri Dec  4 10:29:34 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Fri, 04 Dec 2009 10:29:34 +0000
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B184054.80408@laganowski.net>
References: <4B184054.80408@laganowski.net>
Message-ID: <4B18E48E.9090509@thekelleys.org.uk>

Eric Laganowski wrote:
> I guess I am not communicating this well. The desired dnsmasq
> behavior would be to reply, say, with 192.168.1.1 to any request
> starting with "wpad.", not just local domain, so if, say my domain is
> domain.local, dnsmasq responds with 192.168.1.1 to both
> wpad.domain.local and wpad.google.com
> 
> -Eric

That's not possible. If one wanted to implement it, the logical way 
would be to do full regexp pattern matching on the domains. That has 
been suggested in the past, but I've always resisted it on the grounds 
that it's overkill.

For wpad, the browser will always append a domain, I think. Can you not 
just enumerate all the possible domains?

Agreed that DHCP would be  better way to do this, but also agreed that 
it don't work on firefox (mainly because Linux lacks an API to the DHCP 
system than firefox can use, and the standards for DHCP say you need to 
send DHCPINFORM packets from a privileged port, which makes doing so 
from process run by an ordinary user impossible.)

Simon.


From simon at thekelleys.org.uk  Fri Dec  4 11:17:09 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Fri, 04 Dec 2009 11:17:09 +0000
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <4B184C9F.1040705@gmail.com>
References: <4B184C9F.1040705@gmail.com>
Message-ID: <4B18EFB5.5080204@thekelleys.org.uk>

Tom Metro wrote:
> Is there a dnsmasq directive similar to 'alias' that instead of 
> rewriting IP addresses in the query reply, rewrites the domain in the 
> query? So a query for foo.domain always results in a query for bar.domain?
> 
> This is related to the idea of a CNAME, but the dnsmasq 'cname' 
> directive is restricted to bar.domain being local. I'm looking for 
> something that will work for domains that aren't locally controlled. The 
> 'address' directive is also related, but the target needs to be a 
> domain, or at least a cached lookup, and not a static IP.
> 
>   -Tom
> 

There's no way to do that, and it would be very difficult to provide one 
for the following reason. Dnsmasq doesn't store a query when it forwards 
it: it keeps the minimum amount of information needed to recognise the 
reply and send it back to the original requestor. If the domain was 
re-written before forwarding, there would be no-way to restore the 
original question before returning the answer. The original requestor 
would therefore get an answer to a different question than  the one it 
asked, and ignore it.


Sorry.

Simon.



From redbully at cc.hs-owl.de  Fri Dec  4 11:33:45 2009
From: redbully at cc.hs-owl.de (Jan 'RedBully' Seiffert)
Date: Fri, 04 Dec 2009 12:33:45 +0100
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B18E48E.9090509@thekelleys.org.uk>
References: <4B184054.80408@laganowski.net>
	<4B18E48E.9090509@thekelleys.org.uk>
Message-ID: <4B18F399.7020909@cc.hs-owl.de>

Simon Kelley schrieb:
> Eric Laganowski wrote:
>> I guess I am not communicating this well. The desired dnsmasq
>> behavior would be to reply, say, with 192.168.1.1 to any request
>> starting with "wpad.", not just local domain, so if, say my domain is
>> domain.local, dnsmasq responds with 192.168.1.1 to both
>> wpad.domain.local and wpad.google.com
>>
>> -Eric
> 
> That's not possible. If one wanted to implement it, the logical way 
> would be to do full regexp pattern matching on the domains. That has 
> been suggested in the past, but I've always resisted it on the grounds 
> that it's overkill.
> 

But doesn't mean there is no patch.
I have this laying around for some time, here for dnsmasq 2.50.

This way one can write:
address=/:^wpad\..*:/192.168.0.1
or something like that...

Note: This patch is not that well tested...

[snip]
> 
> Simon.
> 

Greetings
	Jan

-- 
Miksch's Law:
	If a string has one end, then it has another end.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-2.50-regex.patch
Type: text/x-patch
Size: 5577 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091204/e1e6d4c8/attachment-0001.bin 

From richardvoigt at gmail.com  Fri Dec  4 16:03:17 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Fri, 4 Dec 2009 10:03:17 -0600
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
Message-ID: <2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>

On Thu, Dec 3, 2009 at 10:06 PM, Perette Barella <perette at barella.org> wrote:
> I think there's a misunderstanding on how the WPAD DNS version operates. The "wpad.domain.localnet" is used by the browser at startup to locate the proxy configuration file which applies to all domains. ?You don't need a separate wpad.google.com and wpad.amazon.com for every domain users are trying to connect to.
>
> If for some reason your local hosts are configured with different domain names (and therefore looking up wpad.google.com or wpad.amazon.com), I think we need more explanation on just what strangeness you've got going on.

In general, I think we can say that users who have ignored the
DHCP-provided domain and configured their own intend to opt-out of
wpad.  Browser proxy settings are at the discretion of the user
anyway, if you want a mandatory proxy setup you'll need to use
iptables to accomplish that, not DNS.

There's no need to wildcard match wpad hostnames, which are subject to
user-side DNS caching anyway (a user who has configured for
domain=google.com probably already has wpad.google.com cached and
won't get information from dnsmasq).

Any solution to this which involves DNS is inherently broken.

>
> Perette
>
>
> You can do the same thing by
> On 2009?12?03?, at 22:45, Eric Laganowski wrote:
>
>> Well, while legal advice is always appreciated that was not what I was
>> asking for.
>> I was asking about a specific feature of dnsmasq and I am still at a
>> loss whether it is possible to accomplish what I was looking for or not.
>>
>> On a side note, DHCP option 252 is not supported by Firefox, that is way
>> I am forced to explore other options.
>>
>> Thanks,
>> Eric
>>
>> richardvoigt at gmail.com wrote:
>>> I think you're going about this the wrong way. ?Use the DHCP option
>>> "auto-proxy-config" to control the URL browsers use for auto-proxy.
>>> Spoofing addresses in other domains doesn't solve any problems, it
>>> only creates more (and is borderline illegal in many areas).
>>>
>>> On Thu, Dec 3, 2009 at 4:48 PM, Eric Laganowski <eric at laganowski.net> wrote:
>>>
>>>> I guess I am not communicating this well.
>>>> The desired dnsmasq behavior would be to reply, say, with 192.168.1.1 to any request starting with "wpad.", not just local domain, so if, say my domain is domain.local, dnsmasq responds with 192.168.1.1 to both wpad.domain.local and wpad.google.com
>>>>
>>>> -Eric
>>>>
>>>> Date: Wed, 2 Dec 2009 12:52:35 -0430
>>>> From: Santiago Zarate <santiago at zarate.net.ve>
>>>> Subject: Re: [Dnsmasq-discuss] DNS pattern response
>>>> To: dnsmasq-discuss at lists.thekelleys.org.uk
>>>> Message-ID: <200912021252.35536.santiago at zarate.net.ve>
>>>> Content-Type: Text/Plain; ?charset="us-ascii"
>>>>
>>>> i guess you can use a cname record...
>>>> address=/wpad.mydomain.net/10.40.60.90
>>>> cname=wpad,wpad.mydomain.net
>>>>
>>>> tho... i'm pretty sure that cname=wpad,10.40.60.90 will also work...
>>>>
>>>>
>>>> -- Santiago Zarate santiago at zarate.net.ve (+58) 4129864175 (+58) 4241073905
>>>>
>>>>
>>>> Date: Wed, 2 Dec 2009 11:45:57 -0500
>>>> From: Eric Laganowski <elaganowski at hotmail.com>
>>>> Subject: [Dnsmasq-discuss] DNS pattern response
>>>> To: <dnsmasq-discuss at lists.thekelleys.org.uk>
>>>> Message-ID: <SNT130-w394FD18999D0797477ACB7A6950 at phx.gbl>
>>>> Content-Type: text/plain; charset="iso-8859-1"
>>>>
>>>>
>>>> Hi!
>>>>
>>>> I am playing with browser proxy autodiscovery feature and would like dnsmasq to reply with a certain IP address to any DNS query starting with "wpad.", any domain might follow.
>>>> Is it possible to accomplish this with dnsmasq?
>>>>
>>>> -Eric
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eric at laganowski.net  Fri Dec  4 16:27:01 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Fri, 04 Dec 2009 11:27:01 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
Message-ID: <4B193855.2060201@laganowski.net>

richardvoigt at gmail.com wrote:
> On Thu, Dec 3, 2009 at 10:06 PM, Perette Barella <perette at barella.org> wrote:
>   
>> I think there's a misunderstanding on how the WPAD DNS version operates. The "wpad.domain.localnet" is used by the browser at startup to locate the proxy configuration file which applies to all domains.  You don't need a separate wpad.google.com and wpad.amazon.com for every domain users are trying to connect to.
>>
>> If for some reason your local hosts are configured with different domain names (and therefore looking up wpad.google.com or wpad.amazon.com), I think we need more explanation on just what strangeness you've got going on.
>>     
>
> In general, I think we can say that users who have ignored the
> DHCP-provided domain and configured their own intend to opt-out of
> wpad.  Browser proxy settings are at the discretion of the user
> anyway, if you want a mandatory proxy setup you'll need to use
> iptables to accomplish that, not DNS.
>
> There's no need to wildcard match wpad hostnames, which are subject to
> user-side DNS caching anyway (a user who has configured for
> domain=google.com probably already has wpad.google.com cached and
> won't get information from dnsmasq).
>
> Any solution to this which involves DNS is inherently broken.
Guys, all I want to do is to be able to use my company-provided laptop 
at home which has proxy in the network. It is configured with a 
different domain than my local subnet for obvious reasons.
DHCP was tested and confirmed to work properly with MSIE. FF does not 
work as it relies purely on DNS (wpad). The idea is to make this as 
transparent as possible.

-Eric


From eric at laganowski.net  Fri Dec  4 16:29:04 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Fri, 04 Dec 2009 11:29:04 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B18F399.7020909@cc.hs-owl.de>
References: <4B184054.80408@laganowski.net>
	<4B18E48E.9090509@thekelleys.org.uk> <4B18F399.7020909@cc.hs-owl.de>
Message-ID: <4B1938D0.2030502@laganowski.net>

Jan 'RedBully' Seiffert wrote:
> Simon Kelley schrieb:
>   
>> Eric Laganowski wrote:
>>     
>>> I guess I am not communicating this well. The desired dnsmasq
>>> behavior would be to reply, say, with 192.168.1.1 to any request
>>> starting with "wpad.", not just local domain, so if, say my domain is
>>> domain.local, dnsmasq responds with 192.168.1.1 to both
>>> wpad.domain.local and wpad.google.com
>>>
>>> -Eric
>>>       
>> That's not possible. If one wanted to implement it, the logical way 
>> would be to do full regexp pattern matching on the domains. That has 
>> been suggested in the past, but I've always resisted it on the grounds 
>> that it's overkill.
>>
>>     
>
> But doesn't mean there is no patch.
> I have this laying around for some time, here for dnsmasq 2.50.
>
> This way one can write:
> address=/:^wpad\..*:/192.168.0.1
> or something like that...
>   
Jan,

 I appreciate your response. Will test is as soon as possible and 
provide with a feedback.

Thanks,
Eric


From santiago at zarate.net.ve  Fri Dec  4 17:18:27 2009
From: santiago at zarate.net.ve (Santiago Zarate)
Date: Fri, 4 Dec 2009 12:48:27 -0430
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B193855.2060201@laganowski.net>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net>
Message-ID: <a37e9c560912040918p1af6588do194d18b18401df2c@mail.gmail.com>

add an IPTABLES rule.... and that's it..

if i'm not mistaken:

# DNAT port 80 request comming from LAN systems to squid 3128
($SQUID_PORT) aka transparent proxy
iptables -t nat -A PREROUTING -i $LAN_IN -p tcp --dport 80 -j DNAT
--to $SQUID_SERVER:$SQUID_PORT
# if it is same system
iptables -t nat -A PREROUTING -i $INTERNET -p tcp --dport 80 -j
REDIRECT --to-port $SQUID_PORT

Taken from here:
http://www.cyberciti.biz/tips/linux-setup-transparent-proxy-squid-howto.html

2009/12/4 Eric Laganowski <eric at laganowski.net>:
> richardvoigt at gmail.com wrote:
>> On Thu, Dec 3, 2009 at 10:06 PM, Perette Barella <perette at barella.org> wrote:
>>
>>> I think there's a misunderstanding on how the WPAD DNS version operates. The "wpad.domain.localnet" is used by the browser at startup to locate the proxy configuration file which applies to all domains. ?You don't need a separate wpad.google.com and wpad.amazon.com for every domain users are trying to connect to.
>>>
>>> If for some reason your local hosts are configured with different domain names (and therefore looking up wpad.google.com or wpad.amazon.com), I think we need more explanation on just what strangeness you've got going on.
>>>
>>
>> In general, I think we can say that users who have ignored the
>> DHCP-provided domain and configured their own intend to opt-out of
>> wpad. ?Browser proxy settings are at the discretion of the user
>> anyway, if you want a mandatory proxy setup you'll need to use
>> iptables to accomplish that, not DNS.
>>
>> There's no need to wildcard match wpad hostnames, which are subject to
>> user-side DNS caching anyway (a user who has configured for
>> domain=google.com probably already has wpad.google.com cached and
>> won't get information from dnsmasq).
>>
>> Any solution to this which involves DNS is inherently broken.
> Guys, all I want to do is to be able to use my company-provided laptop
> at home which has proxy in the network. It is configured with a
> different domain than my local subnet for obvious reasons.
> DHCP was tested and confirmed to work properly with MSIE. FF does not
> work as it relies purely on DNS (wpad). The idea is to make this as
> transparent as possible.
>
> -Eric
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From b-morgan at concentric.net  Fri Dec  4 17:30:54 2009
From: b-morgan at concentric.net (Brad Morgan)
Date: Fri, 4 Dec 2009 10:30:54 -0700
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B193855.2060201@laganowski.net>
References: <4B184054.80408@laganowski.net>	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>	<4B1885D6.1050105@laganowski.net>	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net>
Message-ID: <003301ca7507$8dae4630$a90ad290$@net>

> Guys, all I want to do is to be able to use my company-provided laptop 
> at home which has proxy in the network. It is configured with a 
> different domain than my local subnet for obvious reasons.
> DHCP was tested and confirmed to work properly with MSIE. FF does not 
> work as it relies purely on DNS (wpad). The idea is to make this as 
> transparent as possible.

I think you need two things. First, an DNS entry on your home network that
resolves wpad.company.network to a local address. Second, at that address,
you need to provide a web server that serves up a proxy configuration file
that basically says bypass the proxy for everything. Here's an example
proxy.pac file:

function FindProxyForURL(url, host)
{
/*
** Proxy configuration file 
**
** Comment out the alert statements by adding // at the
** beginning of the line.
*/
//    alert("url= " + url);
//    alert("host= " + host);
    if (isPlainHostName(host)) {
//        alert("host= " + host + " return= DIRECT (isPlainHostName)");
        return "DIRECT";
    }
    if (isInNet(host, "192.168.0.0", "255.255.255.0")) { // For testing at
home.
//        alert("host= " + host + " return= DIRECT (isInNet)");
        return "DIRECT";
    }
//    alert("host= " + host + " return= PROXY");
    return "PROXY 192.168.0.100:3180"; // Proxy for testing at home
}

 



From eric at laganowski.net  Fri Dec  4 17:40:53 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Fri, 04 Dec 2009 12:40:53 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <a37e9c560912040918p1af6588do194d18b18401df2c@mail.gmail.com>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net>
	<a37e9c560912040918p1af6588do194d18b18401df2c@mail.gmail.com>
Message-ID: <4B1949A5.3000807@laganowski.net>

Santiago Zarate wrote:
> add an IPTABLES rule.... and that's it..
>
> if i'm not mistaken:
>
> # DNAT port 80 request comming from LAN systems to squid 3128
> ($SQUID_PORT) aka transparent proxy
> iptables -t nat -A PREROUTING -i $LAN_IN -p tcp --dport 80 -j DNAT
> --to $SQUID_SERVER:$SQUID_PORT
> # if it is same system
> iptables -t nat -A PREROUTING -i $INTERNET -p tcp --dport 80 -j
> REDIRECT --to-port $SQUID_PORT
>
> Taken from here:
> http://www.cyberciti.biz/tips/linux-setup-transparent-proxy-squid-howto.html
I do really appreciate your help guys. I was trying to be as 
dnsmasq-centric as possible, so some stuff was left out.
Another thing that I am trying to accomplish is to make IWA work and 
SQUID fake_ntlm_auth authenticator working for accounting purposes.
Transparent proxying won't help in this scenario unfortunately.
-Eric


From eric at laganowski.net  Fri Dec  4 17:46:30 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Fri, 04 Dec 2009 12:46:30 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <003301ca7507$8dae4630$a90ad290$@net>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net> <003301ca7507$8dae4630$a90ad290$@net>
Message-ID: <4B194AF6.9060305@laganowski.net>

Brad Morgan wrote:
>> Guys, all I want to do is to be able to use my company-provided laptop 
>> at home which has proxy in the network. It is configured with a 
>> different domain than my local subnet for obvious reasons.
>> DHCP was tested and confirmed to work properly with MSIE. FF does not 
>> work as it relies purely on DNS (wpad). The idea is to make this as 
>> transparent as possible.
>>     
>
> I think you need two things. First, an DNS entry on your home network that
> resolves wpad.company.network to a local address. Second, at that address,
> you need to provide a web server that serves up a proxy configuration file
> that basically says bypass the proxy for everything. Here's an example
> proxy.pac file:
>   
Brad,

 All the pac files have been setup and confirmed working. I can setup 
wpad.company.network and it will work.
I want to accomplish, if you say, a generic kind of design. I do not 
want to be dependent on modifying/restarting dnsmasq
every time a client shows up on my network with a different domain name 
in it's network properties. I understand that
some people may find it questionable/crazy, but I think it is 
technically possible.

Thanks,
Eric


From richardvoigt at gmail.com  Fri Dec  4 17:47:28 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Fri, 4 Dec 2009 11:47:28 -0600
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B193855.2060201@laganowski.net>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net>
Message-ID: <2e59e6970912040947v3b2a0dbfxc79d355aaaea8433@mail.gmail.com>

On Fri, Dec 4, 2009 at 10:27 AM, Eric Laganowski <eric at laganowski.net> wrote:
> richardvoigt at gmail.com wrote:
>>
>> On Thu, Dec 3, 2009 at 10:06 PM, Perette Barella <perette at barella.org>
>> wrote:
>>
>>>
>>> I think there's a misunderstanding on how the WPAD DNS version operates.
>>> The "wpad.domain.localnet" is used by the browser at startup to locate the
>>> proxy configuration file which applies to all domains. ?You don't need a
>>> separate wpad.google.com and wpad.amazon.com for every domain users are
>>> trying to connect to.
>>>
>>> If for some reason your local hosts are configured with different domain
>>> names (and therefore looking up wpad.google.com or wpad.amazon.com), I think
>>> we need more explanation on just what strangeness you've got going on.
>>>
>>
>> In general, I think we can say that users who have ignored the
>> DHCP-provided domain and configured their own intend to opt-out of
>> wpad. ?Browser proxy settings are at the discretion of the user
>> anyway, if you want a mandatory proxy setup you'll need to use
>> iptables to accomplish that, not DNS.
>>
>> There's no need to wildcard match wpad hostnames, which are subject to
>> user-side DNS caching anyway (a user who has configured for
>> domain=google.com probably already has wpad.google.com cached and
>> won't get information from dnsmasq).
>>
>> Any solution to this which involves DNS is inherently broken.
>
> Guys, all I want to do is to be able to use my company-provided laptop at
> home which has proxy in the network. It is configured with a different
> domain than my local subnet for obvious reasons.
> DHCP was tested and confirmed to work properly with MSIE. FF does not work
> as it relies purely on DNS (wpad). The idea is to make this as transparent
> as possible.

And when your laptop has the IP address of wpad.mycompany.com already
in the local cache?  dnsmasq cannot solve this, you need to use
iptables to force traffic through a proxy.  Santiago showed you how to
configure that.

>
> -Eric
>


From tmetro+dnsmasq at gmail.com  Fri Dec  4 17:48:26 2009
From: tmetro+dnsmasq at gmail.com (Tom Metro)
Date: Fri, 04 Dec 2009 12:48:26 -0500
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <4B18EFB5.5080204@thekelleys.org.uk>
References: <4B184C9F.1040705@gmail.com> <4B18EFB5.5080204@thekelleys.org.uk>
Message-ID: <4B194B6A.8030204@gmail.com>

richardvoigt at gmail.com wrote:
> Spoofing addresses in other people's domains doesn't solve any
> problems, it only creates more (and is borderline illegal in many
> areas).

No more illegal than modifying your own hosts file. The scope of the 
modification is, of course, relevant. Doing this at an ISP would be bad. 
Doing this on your private LAN, not a problem.


> What are you trying to accomplish?

The use case is a web server that has both public and private IPs, where 
the private IPs are accessible through a VPN. If VPN connected 
developers wish to access the server via the private interface (which 
enables diagnostics), while using the public host name (to invoke the 
correct virtual host), mapping the private host name to the public host 
name is one way to do this. Avoiding the use of a static IP in that map 
makes sure things don't break if the IT guys change IP addresses.

There are other, and probably better, ways to do this, such as adding a 
host name alias to the public virtual host that corresponds to a private 
IP address, but that would require both code (the host name is used by 
the application) and configuration changes in production.


Simon Kelley wrote:
> There's no way to do that, and it would be very difficult to provide one 
> for the following reason. Dnsmasq doesn't store a query when it forwards 
> it: it keeps the minimum amount of information needed to recognise the 
> reply and send it back to the original requestor.

Yeah, I had a vague recollection of that from a prior discussion on CNAMEs.


> If the domain was re-written before forwarding, there would be no-way
> to restore the original question before returning the answer.

If it was useful enough functionality, it should certainly be doable. It 
would just require a special case.

One way is having a thread make the request and block until a reply or 
timeout happened. Though you'd probably need to limit the quantity of 
such requests you'd process to avoid being DoSed.

Alternatively, could the memory structure you now keep on outstanding 
queries be modified to hold optional fields or a pointer to an optional 
extended data structure? Then you could add in the supplemental 
information for these special case queries without increasing the size 
of the query state information for typical queries.

Theres always the hack approximation. A cron script that periodically 
looks up the target name, and sends a config change to dnsmasq via dbus 
to add an 'address' entry mapping the public host name to the private IP 
address.

  -Tom

-- 
Tom Metro
Venture Logic, Newton, MA, USA
"Enterprise solutions through open source."
Professional Profile: http://tmetro.venturelogic.com/


From richardvoigt at gmail.com  Fri Dec  4 17:49:34 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Fri, 4 Dec 2009 11:49:34 -0600
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B194AF6.9060305@laganowski.net>
References: <4B184054.80408@laganowski.net>
	<2e59e6970912031617g39143f9awc35b598f4eee3b71@mail.gmail.com>
	<4B1885D6.1050105@laganowski.net>
	<A0ABDF84-B54F-444D-9C73-9F0657F20D97@barella.org>
	<2e59e6970912040803u74e9fa2l63a79cf0958013b2@mail.gmail.com>
	<4B193855.2060201@laganowski.net>
	<003301ca7507$8dae4630$a90ad290$@net>
	<4B194AF6.9060305@laganowski.net>
Message-ID: <2e59e6970912040949y37b71a22w7b3ea5e8900b8b02@mail.gmail.com>

On Fri, Dec 4, 2009 at 11:46 AM, Eric Laganowski <eric at laganowski.net> wrote:
> Brad Morgan wrote:
>>>
>>> Guys, all I want to do is to be able to use my company-provided laptop at
>>> home which has proxy in the network. It is configured with a different
>>> domain than my local subnet for obvious reasons.
>>> DHCP was tested and confirmed to work properly with MSIE. FF does not
>>> work as it relies purely on DNS (wpad). The idea is to make this as
>>> transparent as possible.
>>>
>>
>> I think you need two things. First, an DNS entry on your home network that
>> resolves wpad.company.network to a local address. Second, at that address,
>> you need to provide a web server that serves up a proxy configuration file
>> that basically says bypass the proxy for everything. Here's an example
>> proxy.pac file:
>>
>
> Brad,
>
> All the pac files have been setup and confirmed working. I can setup
> wpad.company.network and it will work.
> I want to accomplish, if you say, a generic kind of design. I do not want to
> be dependent on modifying/restarting dnsmasq
> every time a client shows up on my network with a different domain name in
> it's network properties. I understand that
> some people may find it questionable/crazy, but I think it is technically
> possible.

dnsmasq rereads /etc/hosts without any restart required.  But while
responding to wpad.* is technically possible, it doesn't solve the
problem of automatically making clients use your proxy.

>
> Thanks,
> Eric
>


From richardvoigt at gmail.com  Fri Dec  4 21:12:20 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Fri, 4 Dec 2009 15:12:20 -0600
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <4B194B6A.8030204@gmail.com>
References: <4B184C9F.1040705@gmail.com> <4B18EFB5.5080204@thekelleys.org.uk>
	<4B194B6A.8030204@gmail.com>
Message-ID: <2e59e6970912041312v10757369ib104c5707955f0fe@mail.gmail.com>

On Fri, Dec 4, 2009 at 11:48 AM, Tom Metro <tmetro+dnsmasq at gmail.com> wrote:
> richardvoigt at gmail.com wrote:
>> Spoofing addresses in other people's domains doesn't solve any
>> problems, it only creates more (and is borderline illegal in many
>> areas).
>
> No more illegal than modifying your own hosts file. The scope of the
> modification is, of course, relevant. Doing this at an ISP would be bad.
> Doing this on your private LAN, not a problem.
>
>
>> What are you trying to accomplish?
>
> The use case is a web server that has both public and private IPs, where
> the private IPs are accessible through a VPN. If VPN connected
> developers wish to access the server via the private interface (which
> enables diagnostics), while using the public host name (to invoke the
> correct virtual host), mapping the private host name to the public host
> name is one way to do this. Avoiding the use of a static IP in that map
> makes sure things don't break if the IT guys change IP addresses.
>
> There are other, and probably better, ways to do this, such as adding a
> host name alias to the public virtual host that corresponds to a private
> IP address, but that would require both code (the host name is used by
> the application) and configuration changes in production.

Can the diagnostics be enabled conditionally based on the client IP
address, instead of the server IP address binding?  I guess that would
require clients to route traffic for your company's public IPs through
the VPN (and not just private IPs like I guess is done now).

>
>
> Simon Kelley wrote:
>> There's no way to do that, and it would be very difficult to provide one
>> for the following reason. Dnsmasq doesn't store a query when it forwards
>> it: it keeps the minimum amount of information needed to recognise the
>> reply and send it back to the original requestor.
>
> Yeah, I had a vague recollection of that from a prior discussion on CNAMEs.
>
>
>> If the domain was re-written before forwarding, there would be no-way
>> to restore the original question before returning the answer.
>
> If it was useful enough functionality, it should certainly be doable. It
> would just require a special case.
>
> One way is having a thread make the request and block until a reply or
> timeout happened. Though you'd probably need to limit the quantity of
> such requests you'd process to avoid being DoSed.
>
> Alternatively, could the memory structure you now keep on outstanding
> queries be modified to hold optional fields or a pointer to an optional
> extended data structure? Then you could add in the supplemental
> information for these special case queries without increasing the size
> of the query state information for typical queries.
>
> Theres always the hack approximation. A cron script that periodically
> looks up the target name, and sends a config change to dnsmasq via dbus
> to add an 'address' entry mapping the public host name to the private IP
> address.
>
> ?-Tom
>
> --
> Tom Metro
> Venture Logic, Newton, MA, USA
> "Enterprise solutions through open source."
> Professional Profile: http://tmetro.venturelogic.com/
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From eric at laganowski.net  Fri Dec  4 21:22:53 2009
From: eric at laganowski.net (Eric Laganowski)
Date: Fri, 04 Dec 2009 16:22:53 -0500
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B18F399.7020909@cc.hs-owl.de>
References: <4B184054.80408@laganowski.net>
	<4B18E48E.9090509@thekelleys.org.uk> <4B18F399.7020909@cc.hs-owl.de>
Message-ID: <4B197DAD.1090603@laganowski.net>

Jan 'RedBully' Seiffert wrote:
> I have this laying around for some time, here for dnsmasq 2.50.
>
> This way one can write:
> address=/:^wpad\..*:/192.168.0.1
> or something like that...
>
> Note: This patch is not that well tested...
>   
Had to modify a little. Preliminary tests show expected behavior.

Thanks so much,
Eric

>> Simon.
>>
>>     
>
> Greetings
> 	Jan
>   



From v_cadet at yahoo.fr  Fri Dec  4 22:16:53 2009
From: v_cadet at yahoo.fr (Vincent Cadet)
Date: Fri, 4 Dec 2009 22:16:53 +0000 (GMT)
Subject: [Dnsmasq-discuss] DNS pattern response
In-Reply-To: <4B193855.2060201@laganowski.net>
Message-ID: <811050.80316.qm@web28207.mail.ukl.yahoo.com>

> Guys, all I want to do is to be able to use my
> company-provided laptop 
> at home which has proxy in the network. It is configured
> with a 
> different domain than my local subnet for obvious reasons.
> DHCP was tested and confirmed to work properly with MSIE.
> FF does not 
> work as it relies purely on DNS (wpad). The idea is to make
> this as 
> transparent as possible.

What about using profiles with Firefox? Different profiles, different network/proxy settings. KISS.

If you have a GNU/Linux machine you could even hard link some of your important files (like password file, extensions, bookmarks, form data) across profiles to avoid copying them.

Vince C.


      


From tmetro+dnsmasq at gmail.com  Fri Dec  4 23:04:00 2009
From: tmetro+dnsmasq at gmail.com (Tom Metro)
Date: Fri, 04 Dec 2009 18:04:00 -0500
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <2e59e6970912041312v10757369ib104c5707955f0fe@mail.gmail.com>
References: <4B184C9F.1040705@gmail.com> <4B18EFB5.5080204@thekelleys.org.uk>	
	<4B194B6A.8030204@gmail.com>
	<2e59e6970912041312v10757369ib104c5707955f0fe@mail.gmail.com>
Message-ID: <4B199560.60602@gmail.com>

richardvoigt at gmail.com wrote:
> Tom Metro wrote:
>> The use case is a web server that has both public and private IPs, where
>> the private IPs are accessible through a VPN. If VPN connected
>> developers wish to access the server via the private interface (which
>> enables diagnostics), while using the public host name (to invoke the
>> correct virtual host), mapping the private host name to the public host
>> name is one way to do this.
> 
> Can the diagnostics be enabled conditionally based on the client IP
> address, instead of the server IP address binding?

That's actually what is being done.


> I guess that would require clients to route traffic for your
> company's public IPs through the VPN (and not just private IPs like I
> guess is done now).

Correct, and that is what is accomplished by accessing the server 
through its private IP.

It also works to add custom routes on the client machines to force the 
public IP to go through the VPN, but that also suffers from a potential 
problem with stale IP addresses if the public IP changes. (Not a huge 
deal as the route could be added via a script when the VPN is brought 
up. The script can fetch the current IP, and VPN sessions are relatively 
short lived compared to the frequency that IP addresses change.)

  -Tom

-- 
Tom Metro
Venture Logic, Newton, MA, USA
"Enterprise solutions through open source."
Professional Profile: http://tmetro.venturelogic.com/


From richardvoigt at gmail.com  Sat Dec  5 03:32:35 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Fri, 4 Dec 2009 21:32:35 -0600
Subject: [Dnsmasq-discuss] query rewriting
In-Reply-To: <4B199560.60602@gmail.com>
References: <4B184C9F.1040705@gmail.com> <4B18EFB5.5080204@thekelleys.org.uk>
	<4B194B6A.8030204@gmail.com>
	<2e59e6970912041312v10757369ib104c5707955f0fe@mail.gmail.com>
	<4B199560.60602@gmail.com>
Message-ID: <2e59e6970912041932o511c2a5bha59c69e90f8790d@mail.gmail.com>

On Fri, Dec 4, 2009 at 5:04 PM, Tom Metro <tmetro+dnsmasq at gmail.com> wrote:
> richardvoigt at gmail.com wrote:
>>
>> Tom Metro wrote:
>>>
>>> The use case is a web server that has both public and private IPs, where
>>> the private IPs are accessible through a VPN. If VPN connected
>>> developers wish to access the server via the private interface (which
>>> enables diagnostics), while using the public host name (to invoke the
>>> correct virtual host), mapping the private host name to the public host
>>> name is one way to do this.
>>
>> Can the diagnostics be enabled conditionally based on the client IP
>> address, instead of the server IP address binding?
>
> That's actually what is being done.
>
>
>> I guess that would require clients to route traffic for your
>> company's public IPs through the VPN (and not just private IPs like I
>> guess is done now).
>
> Correct, and that is what is accomplished by accessing the server through
> its private IP.
>
> It also works to add custom routes on the client machines to force the
> public IP to go through the VPN, but that also suffers from a potential
> problem with stale IP addresses if the public IP changes. (Not a huge deal
> as the route could be added via a script when the VPN is brought up. The
> script can fetch the current IP, and VPN sessions are relatively short lived
> compared to the frequency that IP addresses change.)

Any downside to routing the entire IP block for your company into the
VPN?  I presume that changes even less frequently than server
addresses.


From clintonlee.taylor at gmail.com  Sun Dec 13 15:41:11 2009
From: clintonlee.taylor at gmail.com (Clinton Lee Taylor)
Date: Sun, 13 Dec 2009 17:41:11 +0200
Subject: [Dnsmasq-discuss] Multiple resolv.conf ...
Message-ID: <b7b14cbb0912130741p3733f3fakcc17e2d7ef47bab5@mail.gmail.com>

Greetings ...

I know we have discussed mltiple resolv.conf before, but I think that
either I need to ask for a new feature, or that we might have a bug
...

 This weekend, I was trying to setup a 3G USB modem as a backup
internet connection for a server.  It seems that pppd has a bug, that
get's faulty DNS from the RAS.  This would not be a problem, if all
the DNS servers I have listed in my multiple resolvers ...

 Let me explain, I have a resolv.conf-google, resolv.conf-opendns,
resolv.conf-local and resolv.conf-pppd.  When pppd connects,
if-up.local runs, which I then copy the DNS1 and DNS2 into
resolv.conf-pppd.

 Trying to ping www.opndns.com, would not work, but ping an IP did ...

Then when pppd disconnects, I run if-down.local, which copies
resolv.conf-local over resolv.conf-pppd and pings work again ...

 I also restart dnsmasq in ip-up.local and ip-down.local, and only see
the last file resolver file and it's DNS.

Could we please list all DNS servers when starting dnsmasq, not just
the last updated DNS?

Thanks
Mailed
LeeT


From bcook at poughkeepsieschools.org  Mon Dec 14 20:26:42 2009
From: bcook at poughkeepsieschools.org (B. Cook)
Date: Mon, 14 Dec 2009 15:26:42 -0500
Subject: [Dnsmasq-discuss] why sending to resolv-file server?
Message-ID: <4B269F82.10801@poughkeepsieschools.org>

Hello all,

Dnsmasq version 2.50  Copyright (C) 2000-2009 Simon Kelley
Compile time options no-IPv6 GNU-getopt no-DBus no-I18N DHCP TFTP


I have a rather large setup (some 2500+ hosts) all getting dhcp from 
dnsmasq.

I have 10 buildings all setup with dhcp-relay and such to get them back 
to the main one.

My question is:

I do not understand why requests are getting sent to the recursor when 
they do not exist..

so say for example I have the internal domain name .pcsd and each 
building has a subdomain bldg1.pcsd, bldg2.pcsd..

(each building has its own config) and we block unknowns from requesting 
dhcp so we have a large file with all the dhcp-host entries in it that 
each building config includes..

Here is a basic entry from one of the files..

domain=war.pcsd,10.20.24.0/22
dhcp-range=war,10.20.24.33,10.20.25.128,255.255.252.0,4h
dhcp-option=war,option:router,10.20.24.1

and if dnsmasq serves the domain war.pcsd I do not need to put in an 
entry like this:

server=/war.pcsd/10.20.0.16
server=/24.20.10.in-addr.arpa/10.20.0.16

yet, with the entry in or out if I do this:

host totallyfakename.war.pcsd

and then check the query.log for the recursor (bind in this case) it 
will show me that bind was asked for the entry.  Even though I directly 
asked dnsmasq.

Is this intended?

Or do I have something in one of my configs that is helping me shoot 
myself in the foot..

:)

Thanks in advance.



From richardvoigt at gmail.com  Mon Dec 14 22:54:57 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Mon, 14 Dec 2009 16:54:57 -0600
Subject: [Dnsmasq-discuss] why sending to resolv-file server?
In-Reply-To: <4B269F82.10801@poughkeepsieschools.org>
References: <4B269F82.10801@poughkeepsieschools.org>
Message-ID: <2e59e6970912141454v4db70b3fo826f011b3075758f@mail.gmail.com>

On Mon, Dec 14, 2009 at 2:26 PM, B. Cook <bcook at poughkeepsieschools.org> wrote:
> Hello all,
>
> Dnsmasq version 2.50 ?Copyright (C) 2000-2009 Simon Kelley
> Compile time options no-IPv6 GNU-getopt no-DBus no-I18N DHCP TFTP
>
>
> I have a rather large setup (some 2500+ hosts) all getting dhcp from
> dnsmasq.
>
> I have 10 buildings all setup with dhcp-relay and such to get them back
> to the main one.
>
> My question is:
>
> I do not understand why requests are getting sent to the recursor when
> they do not exist..
>
> so say for example I have the internal domain name .pcsd and each
> building has a subdomain bldg1.pcsd, bldg2.pcsd..
>
> (each building has its own config) and we block unknowns from requesting
> dhcp so we have a large file with all the dhcp-host entries in it that
> each building config includes..
>
> Here is a basic entry from one of the files..
>
> domain=war.pcsd,10.20.24.0/22
> dhcp-range=war,10.20.24.33,10.20.25.128,255.255.252.0,4h
> dhcp-option=war,option:router,10.20.24.1
>
> and if dnsmasq serves the domain war.pcsd I do not need to put in an
> entry like this:
>
> server=/war.pcsd/10.20.0.16
> server=/24.20.10.in-addr.arpa/10.20.0.16
>
> yet, with the entry in or out if I do this:
>
> host totallyfakename.war.pcsd
>
> and then check the query.log for the recursor (bind in this case) it
> will show me that bind was asked for the entry. ?Even though I directly
> asked dnsmasq.


From luca.postregna at gmail.com  Thu Dec 17 21:51:08 2009
From: luca.postregna at gmail.com (Luca Postregna)
Date: Thu, 17 Dec 2009 22:51:08 +0100
Subject: [Dnsmasq-discuss] unable to resolv hosts by name in my local network
Message-ID: <d5680d110912171351i6a1c750eo606b9701edf35e62@mail.gmail.com>

Is there a way to resolv by name the hosts of my network without modify
/etc/hosts on the server or in the client?
On the server in dhcp.leased I can read that my client send the hostname, is
not that enough?

Thanks, Luca.



-- 
http://www.infis.units.it/~lucapost/
Luca Postregna
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091217/0775c4a4/attachment.htm 

From ranceh at gmail.com  Thu Dec 17 22:02:27 2009
From: ranceh at gmail.com (Rance Hall)
Date: Thu, 17 Dec 2009 16:02:27 -0600
Subject: [Dnsmasq-discuss] unable to resolv hosts by name in my local
	network
In-Reply-To: <d5680d110912171351i6a1c750eo606b9701edf35e62@mail.gmail.com>
References: <d5680d110912171351i6a1c750eo606b9701edf35e62@mail.gmail.com>
Message-ID: <845c0f80912171402r3ed2f6f0t478d7010488a0dc5@mail.gmail.com>

On Thu, Dec 17, 2009 at 3:51 PM, Luca Postregna
<luca.postregna at gmail.com> wrote:
> Is there a way to resolv by name the hosts of my network without modify
> /etc/hosts on the server or in the client?
> On the server in dhcp.leased I can read that my client send the hostname, is
> not that enough?
>
> Thanks, Luca.
>
>
> --
> http://www.infis.units.it/~lucapost/
> Luca Postregna

The clients that get dhcp addresses should already be able to resolv
by name the other dhcp clients.  If this is not true, you have a
problem and you need to fix that first.

In order for the clients to resolv by name static ip machines, the
server needs to know about them some way.

typically this is with the /etc/hosts file, but IIRC dnsmasq has the
ability to specify an additional/alternate /etc/hosts file.  this
other file needs to be a properly formatted /etc/hosts file, it can
just have another name.

This should get your clients resolving the network correctly.

The next thing you need to know about is if you want the dhcp server
to resolv local names the same way the clients do.

There is a way to configure this also.

specify in your servers /etc/resolv.conf that the server should look
at itself for name resolution, and then specify a second proper
/etc/resolv.conf file for dnsmasq to use to get the rest of the net.


From luca.postregna at gmail.com  Fri Dec 18 09:31:59 2009
From: luca.postregna at gmail.com (Luca Postregna)
Date: Fri, 18 Dec 2009 10:31:59 +0100
Subject: [Dnsmasq-discuss] unable to resolv hosts by name in my local
	network
In-Reply-To: <845c0f80912171402r3ed2f6f0t478d7010488a0dc5@mail.gmail.com>
References: <d5680d110912171351i6a1c750eo606b9701edf35e62@mail.gmail.com>
	<845c0f80912171402r3ed2f6f0t478d7010488a0dc5@mail.gmail.com>
Message-ID: <d5680d110912180131l26fd7bc6q203f8bd2f5589d84@mail.gmail.com>

Ok, this is true. But my question is different, i think.
If in /etc/hosts are present my hosts, with expandhost option in my dnsmasq
config, resolv by name work properly.
But I don't want to specify anything in /etc/hosts. I think that dnsmasq,
when one client get address, obtain all info (dhcp.leases) to permit itself
and other clients to resolv by name the new client, is it true?

LP



On Thu, Dec 17, 2009 at 11:02 PM, Rance Hall <ranceh at gmail.com> wrote:

> On Thu, Dec 17, 2009 at 3:51 PM, Luca Postregna
> <luca.postregna at gmail.com> wrote:
> > Is there a way to resolv by name the hosts of my network without modify
> > /etc/hosts on the server or in the client?
> > On the server in dhcp.leased I can read that my client send the hostname,
> is
> > not that enough?
> >
> > Thanks, Luca.
> >
> >
> > --
> > http://www.infis.units.it/~lucapost/<http://www.infis.units.it/%7Elucapost/>
> > Luca Postregna
>
> The clients that get dhcp addresses should already be able to resolv
> by name the other dhcp clients.  If this is not true, you have a
> problem and you need to fix that first.
>
> In order for the clients to resolv by name static ip machines, the
> server needs to know about them some way.
>
> typically this is with the /etc/hosts file, but IIRC dnsmasq has the
> ability to specify an additional/alternate /etc/hosts file.  this
> other file needs to be a properly formatted /etc/hosts file, it can
> just have another name.
>
> This should get your clients resolving the network correctly.
>
> The next thing you need to know about is if you want the dhcp server
> to resolv local names the same way the clients do.
>
> There is a way to configure this also.
>
> specify in your servers /etc/resolv.conf that the server should look
> at itself for name resolution, and then specify a second proper
> /etc/resolv.conf file for dnsmasq to use to get the rest of the net.
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>



-- 
http://www.infis.units.it/~lucapost/
Luca Postregna
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091218/17b58ea9/attachment.htm 

From ranceh at gmail.com  Fri Dec 18 16:36:12 2009
From: ranceh at gmail.com (Rance Hall)
Date: Fri, 18 Dec 2009 10:36:12 -0600
Subject: [Dnsmasq-discuss] unable to resolv hosts by name in my local
	network
In-Reply-To: <d5680d110912180131l26fd7bc6q203f8bd2f5589d84@mail.gmail.com>
References: <d5680d110912171351i6a1c750eo606b9701edf35e62@mail.gmail.com>
	<845c0f80912171402r3ed2f6f0t478d7010488a0dc5@mail.gmail.com>
	<d5680d110912180131l26fd7bc6q203f8bd2f5589d84@mail.gmail.com>
Message-ID: <845c0f80912180836k60fc680fk8351de6000ac6115@mail.gmail.com>

On Fri, Dec 18, 2009 at 3:31 AM, Luca Postregna
<luca.postregna at gmail.com> wrote:
> Ok, this is true. But my question is different, i think.
> If in /etc/hosts are present my hosts, with expandhost option in my dnsmasq
> config, resolv by name work properly.
> But I don't want to specify anything in /etc/hosts. I think that dnsmasq,
> when one client get address, obtain all info (dhcp.leases) to permit itself
> and other clients to resolv by name the new client, is it true?
>
> LP


<snip>

This is almost always true for cases when all clients that need to
resolv each others names are provided by the dhcp server.

You don't have to modify /etc/hosts if dhcp clients are all you have.


There are some common things that interfere with this basic status.

1) Some OSes (usually unix based) ship needing dhclient to be
configured to send a hostname with the dhcp request.  Windows does
this by default, but with *nix this is a configurable option.  NOT
sending the hostname with the dhcp request will cause name resolution
to fail for just these hosts.


2) Its possible (though I've only seen it happen once in my several
years of using dnsmasq) to get a corrupted dhcp.leases file.  If this
file is hosed then name resolution of dhcp clients can be broken
depending on the specific way the leases file gets hosed.

3) the wrong "authoritative" setting in dnsmasq.conf or another dhcp
server somewhere on your net that also thinks its authoritative.

There might be other "off the wall" things as well.

I refer you to my previous post, the behavior you desire is in my
view, default.  If you don't get it, there is something broken.  But
you need to do a better job of explaining what "broken" looks like for
you so we can help you figure out what to do about it.


From sfzhi at yahoo.com  Sun Dec 20 14:21:58 2009
From: sfzhi at yahoo.com (Sergei Zhirikov)
Date: Sun, 20 Dec 2009 15:21:58 +0100
Subject: [Dnsmasq-discuss] Is this a documentation error?
Message-ID: <hglbuh$kt7$1@ger.gmane.org>

Hi,

The man page of dnsmasq-2.51 contains the following line:

-F, --dhcp-range=[[net:]network-id,]<start-addr>,<end-addr>[[,<netmask>],<broadcast>][,<lease time>]

Looking at the options parsing code I would say it should be:

-F, --dhcp-range=[[net:]network-id,]<start-addr>,<end-addr>[,<netmask>[,<broadcast>]][,<lease time>]

The latter also appears to match the examples in dnsmasq.conf.example.
Is that correct or did I misunderstand something?

--
Thanks and kind regards,
Sergei.



From bretschneider-mldnsmasq at imr.uni-hannover.de  Mon Dec 21 10:55:58 2009
From: bretschneider-mldnsmasq at imr.uni-hannover.de (Martin Bretschneider)
Date: Mon, 21 Dec 2009 11:55:58 +0100
Subject: [Dnsmasq-discuss] exception of an dns entry of the local net
Message-ID: <4B2F543E.4090104@imr.uni-hannover.de>

Hi,

a server with www.localnet is outside my control of localnet. Dnsmasq 
thinks that it is local due to the domain and wants to find it in 
/etc/hosts but it should not be there: I want dnsmasq to ask the default 
DNS server only for www.localnet. Is there an option that www.localnet 
is *not* in the localnet despite the fact that there is the line 
"local=/localnet/" in the config?

TIA  Martin


From simon at thekelleys.org.uk  Mon Dec 21 12:32:53 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 21 Dec 2009 12:32:53 +0000
Subject: [Dnsmasq-discuss] Is this a documentation error?
In-Reply-To: <hglbuh$kt7$1@ger.gmane.org>
References: <hglbuh$kt7$1@ger.gmane.org>
Message-ID: <4B2F6AF5.8050102@thekelleys.org.uk>

Sergei Zhirikov wrote:
> Hi,
> 
> The man page of dnsmasq-2.51 contains the following line:
> 
> -F, --dhcp-range=[[net:]network-id,]<start-addr>,<end-addr>[[,<netmask>],<broadcast>][,<lease time>]
> 
> Looking at the options parsing code I would say it should be:
> 
> -F, --dhcp-range=[[net:]network-id,]<start-addr>,<end-addr>[,<netmask>[,<broadcast>]][,<lease time>]
> 
> The latter also appears to match the examples in dnsmasq.conf.example.
> Is that correct or did I misunderstand something?
> 
You're right: it's a documentation error. I'll fix it in the next release.

Many thanks.

Cheers,

Simon.


From simon at thekelleys.org.uk  Mon Dec 21 16:10:56 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 21 Dec 2009 16:10:56 +0000
Subject: [Dnsmasq-discuss] exception of an dns entry of the local net
In-Reply-To: <4B2F543E.4090104@imr.uni-hannover.de>
References: <4B2F543E.4090104@imr.uni-hannover.de>
Message-ID: <4B2F9E10.9050203@thekelleys.org.uk>

Martin Bretschneider wrote:
> Hi,
> 
> a server with www.localnet is outside my control of localnet. Dnsmasq 
> thinks that it is local due to the domain and wants to find it in 
> /etc/hosts but it should not be there: I want dnsmasq to ask the default 
> DNS server only for www.localnet. Is there an option that www.localnet 
> is *not* in the localnet despite the fact that there is the line 
> "local=/localnet/" in the config?
> 
> TIA  Martin
> 

There's not way to make it ask the default server, but you can make it
ask a server just for www.localnet, and that server could be the same as
the default one.

server=/www.localnet/<server-ip>


HTH

Simon


From bretschneider-mldnsmasq at imr.uni-hannover.de  Mon Dec 21 16:49:15 2009
From: bretschneider-mldnsmasq at imr.uni-hannover.de (Martin Bretschneider)
Date: Mon, 21 Dec 2009 17:49:15 +0100
Subject: [Dnsmasq-discuss] exception of an dns entry of the local net
In-Reply-To: <4B2F9E10.9050203@thekelleys.org.uk>
References: <4B2F543E.4090104@imr.uni-hannover.de>
	<4B2F9E10.9050203@thekelleys.org.uk>
Message-ID: <4B2FA70B.8010907@imr.uni-hannover.de>

Am 21.12.2009 17:10, schrieb Simon Kelley:

Hi Simon,

> Martin Bretschneider wrote:
>> Hi,
>>
>> a server with www.localnet is outside my control of localnet. Dnsmasq
>> thinks that it is local due to the domain and wants to find it in
>> /etc/hosts but it should not be there: I want dnsmasq to ask the default
>> DNS server only for www.localnet. Is there an option that www.localnet
>> is *not* in the localnet despite the fact that there is the line
>> "local=/localnet/" in the config?
>>
>> TIA  Martin
>>
>
> There's not way to make it ask the default server, but you can make it
> ask a server just for www.localnet, and that server could be the same as
> the default one.
>
> server=/www.localnet/<server-ip>

thanks for your anser but Barella gave me PM the tip just to uncomment 
the local domain with
	# local=/localnet/
and it works.

Thanks very much.

Martin


From redbully at cc.hs-owl.de  Mon Dec 21 19:32:09 2009
From: redbully at cc.hs-owl.de (Jan 'RedBully' Seiffert)
Date: Mon, 21 Dec 2009 20:32:09 +0100
Subject: [Dnsmasq-discuss] how to add regex matching ?
In-Reply-To: <2053774952@web.de>
References: <2053774952@web.de>
Message-ID: <4B2FCD39.90309@cc.hs-owl.de>

pulleman at web.de schrieb:
> hi,
> gibt es deinen patch auch f?r die 2.51 ?
> 
> bin kein coder....
> 
> :~/dnsmasq-2.51# patch  < attachment-0001.bin
> patching file Makefile
> Hunk #1 FAILED at 27.
> Hunk #2 FAILED at 36.
> Hunk #3 FAILED at 56.
> 3 out of 3 hunks FAILED -- saving rejects to file Makefile.rej
> patching file Makefile
> Hunk #1 FAILED at 8.
> 1 out of 1 hunk FAILED -- saving rejects to file Makefile.rej
> can't find file to patch at input line 45
> Perhaps you should have used the -p or --strip option?
> The text leading up to this was:
> --------------------------
> |
> |=== modified file 'src/dnsmasq.h'
> |--- old/src/dnsmasq.h  2009-09-10 23:32:18 +0000
> |+++ new/src/dnsmasq.h  2009-09-10 23:41:19 +0000
> --------------------------

Looks, my regex patch does not apply to 2.51, so i imported 2.51 into my VCS,
merged it and regenerated the patch.

Apply with:
cd dnsmasq-2.51
patch -p1 < patchfile_name.patch

Only compile tested.
No warranty it won't eat your kitten.

[snip]
>>>> <hr>
>>>> hi all,
>>>>
>>>>
>>>> running 2.5.1 I would like to reply with 0.0.0.0 for something like ^ad\.*
>>>> I found a hint in early 2007 that a patch exists, unfortunately I can't find it.
>>>> any pointers appreciated.
>>>>
>>>> thanks
>>>> pulle
[snip]

Greetings
	Jan

-- 
assert(!"The excrement has collided with the air circulation device");
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-2.51-regex.patch
Type: text/x-patch
Size: 4708 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091221/56845042/attachment.bin 

From redbully at cc.hs-owl.de  Mon Dec 21 22:05:13 2009
From: redbully at cc.hs-owl.de (Jan 'RedBully' Seiffert)
Date: Mon, 21 Dec 2009 23:05:13 +0100
Subject: [Dnsmasq-discuss] how to add regex matching ?
In-Reply-To: <4B2FCD39.90309@cc.hs-owl.de>
References: <2053774952@web.de> <4B2FCD39.90309@cc.hs-owl.de>
Message-ID: <4B2FF119.4090900@cc.hs-owl.de>

Jan 'RedBully' Seiffert schrieb:
[snip]
> 
> Looks, my regex patch does not apply to 2.51, so i imported 2.51 into my VCS,
> merged it and regenerated the patch.
> 

And while at it, rediff against dnsmasq-2.52test12

> Apply with:
> cd dnsmasq-2.52test12
> patch -p1 < patchfile_name.patch
> 
> Only compile tested.
> No warranty it won't eat your kitten.
> 
Greetings
	Jan

-- 
Networking? That is for fishermen.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dnsmasq-2.52test12-regex.patch
Type: text/x-patch
Size: 4821 bytes
Desc: not available
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091221/cba29fc7/attachment.bin 

From andy at r210.com  Wed Dec 23 19:18:17 2009
From: andy at r210.com (Andy Leak)
Date: Wed, 23 Dec 2009 11:18:17 -0800
Subject: [Dnsmasq-discuss] Not Resolving DHCP addresses
Message-ID: <4cb868300912231118w529c050cy714be71c29b1f20f@mail.gmail.com>

I've got a new DNSMASQ installation.
- DHCP assigns addresses fine
- DNS resolution for external and fixed addresses is fine
- BUT - DNS resolution for DHCP addresses is not working

Here's my dnsmasq.conf file...
domain-needed
domain=x
expand-hosts
resolv-file=/etc/resolv.conf.dnsmasq
dhcp-range=192.168.4.100,192.168.4.200,24h
local=/x/
log-queries
no-negcache
dhcp-authoritative

when I do "nslookup xxx" (where xxx is the name of a dhcp machine) i get

** server can't find xxx.x: NXDOMAIN

when i look at the dnsmasq log, and the dnsmasq.leases file, everything
seems to be ok...

Hmm - how can i get DNSMASQ to resolve my DHCP machines ??
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091223/1cc70840/attachment.htm 

From simon at thekelleys.org.uk  Wed Dec 23 21:19:25 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Wed, 23 Dec 2009 21:19:25 +0000
Subject: [Dnsmasq-discuss] Questio on DHCP option 125
In-Reply-To: <8ade8a010911040441v5f1c826cw398cca193cddf01f@mail.gmail.com>
References: <8ade8a010911040441v5f1c826cw398cca193cddf01f@mail.gmail.com>
Message-ID: <4B32895D.6050701@thekelleys.org.uk>

Bo Zhou wrote:
> Hi,
> 
> I'm a new comer to this mail list. I noticed that 2 years ago, there were
> some people talking about DHCP option 125 support in the dnsmasq. (
> http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2007q3/001590.html)
> I am just wondering what is the status for that. I checked dnsmasq v2.51,
> seems it still doesn't support option 125. Anyone has news on that?
> 
> I do apologize that if there's maill thread already discussed similiar
> question but could anybody give me the link as reference if there is. Thanks
> a lot.
> 
> /Bram
> 
> 

http://www.thekelleys.org.uk/dnsmasq/test-releases/dnsmasq-2.52test15.tar.gz

now includes experimental support for options 124 and 125. Feedback is
solicited.


Cheers,

Simon.


From andy at r210.com  Thu Dec 24 00:01:19 2009
From: andy at r210.com (Andy Leak)
Date: Wed, 23 Dec 2009 16:01:19 -0800
Subject: [Dnsmasq-discuss] Not Resolving DHCP addresses
In-Reply-To: <4cb868300912231118w529c050cy714be71c29b1f20f@mail.gmail.com>
References: <4cb868300912231118w529c050cy714be71c29b1f20f@mail.gmail.com>
Message-ID: <4cb868300912231601q3d152984vc1853183ee79e53a@mail.gmail.com>

Dear DNSMASQ List:

Now DHCP addresses are being resolved just fine, and everything is working
100% ok.

Here's what I did:
- went to lunch
- did a little christmas shopping
- then came back and everything was working fine

Now I am curious why this problem occurred, and what happened to resolve the
issue.  Any ideas??

Thanks for DNSMASQ - it is really a great program.


I've got a new DNSMASQ installation.
> - DHCP assigns addresses fine
> - DNS resolution for external and fixed addresses is fine
> - BUT - DNS resolution for DHCP addresses is not working
>
> Here's my dnsmasq.conf file...
> domain-needed
> domain=x
> expand-hosts
> resolv-file=/etc/resolv.conf.dnsmasq
> dhcp-range=192.168.4.100,192.168.4.200,24h
> local=/x/
> log-queries
> no-negcache
> dhcp-authoritative
>
> when I do "nslookup xxx" (where xxx is the name of a dhcp machine) i get
>
> ** server can't find xxx.x: NXDOMAIN
>
> when i look at the dnsmasq log, and the dnsmasq.leases file, everything
> seems to be ok...
>
> Hmm - how can i get DNSMASQ to resolve my DHCP machines ??
>



-- 
andy at r210.com
650-823-0836
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091223/c6e6fe0c/attachment.htm 

From rob0 at gmx.co.uk  Thu Dec 24 00:37:24 2009
From: rob0 at gmx.co.uk (/dev/rob0)
Date: Wed, 23 Dec 2009 18:37:24 -0600
Subject: [Dnsmasq-discuss] Not Resolving DHCP addresses
In-Reply-To: <4cb868300912231601q3d152984vc1853183ee79e53a@mail.gmail.com>
References: <4cb868300912231118w529c050cy714be71c29b1f20f@mail.gmail.com>
	<4cb868300912231601q3d152984vc1853183ee79e53a@mail.gmail.com>
Message-ID: <20091224003723.GB16229@minipax>

On Wed, Dec 23, 2009 at 04:01:19PM -0800, Andy Leak wrote:
> Now DHCP addresses are being resolved just fine, and everything
> is working 100% ok.
> 
> Here's what I did:
> - went to lunch
> - did a little christmas shopping
> - then came back and everything was working fine
> 
> Now I am curious why this problem occurred, and what happened to
> resolve the issue.  Any ideas??

Very simple: dnsmasq(8) can only serve names that are in its lists,
either hosts(5) or the leases file. If you start up a new dnsmasq on
an existing network, it has no magical means of knowing what leases
the previous DHCP daemon gave out. Over time, leases expired and your
hosts came in to renew them. Your dnsmasq didn't know these were
renewals.

> Thanks for DNSMASQ - it is really a great program.

Your shopping should have included a goodie for Simon! ;)

[the OP]
> I've got a new DNSMASQ installation.
> > - DHCP assigns addresses fine
> > - DNS resolution for external and fixed addresses is fine
> > - BUT - DNS resolution for DHCP addresses is not working
> >
> > Here's my dnsmasq.conf file...
> > domain-needed
> > domain=x
> > expand-hosts
> > resolv-file=/etc/resolv.conf.dnsmasq
> > dhcp-range=192.168.4.100,192.168.4.200,24h
> > local=/x/
> > log-queries
> > no-negcache
> > dhcp-authoritative
> >
> > when I do "nslookup xxx" (where xxx is the name of a dhcp 
> > machine) i get

nslookup(1) is buggy and not a good tool. I recommend that you become
familiar with host(1) and dig(1). All three are provided by BIND. The
ISC folks have threatened to drop nslookup.

> > ** server can't find xxx.x: NXDOMAIN
> >
> > when i look at the dnsmasq log, and the dnsmasq.leases file,
> > everything seems to be ok...

You looked, and at the time, "xxx" was in the lease file? I bet it
wasn't.
-- 
    Offlist mail to this address is discarded unless
    "/dev/rob0" or "not-spam" is in Subject: header


From andy at r210.com  Thu Dec 24 01:54:31 2009
From: andy at r210.com (Andy Leak)
Date: Wed, 23 Dec 2009 17:54:31 -0800
Subject: [Dnsmasq-discuss] Not Resolving DHCP addresses
In-Reply-To: <20091224003723.GB16229@minipax>
References: <4cb868300912231118w529c050cy714be71c29b1f20f@mail.gmail.com>
	<4cb868300912231601q3d152984vc1853183ee79e53a@mail.gmail.com>
	<20091224003723.GB16229@minipax>
Message-ID: <4cb868300912231754l5c92c3b7jd7de1dbe3ea1243@mail.gmail.com>

Thx for your reply...

>nslookup(1) is buggy and not a good tool. I recommend that you become
> >familiar with host(1) and dig(1). All three are provided by BIND. The
> >ISC folks have threatened to drop nslookup.
>

Thanks for that tip.  I'll study host and dig...


> >You looked, and at the time, "xxx" was in the lease file? I bet it
> >wasn't.
>

No - it was in the lease file - I'm sure of it.  And I watched the dnsmasq
log file as the addresses were handed out.  And the addresses in the lease
file matched the address I saw when running 'ifconfig' on the console.

In the course of debugging, I re-assigned the DHCP address over and over.
(by running 'sudo /etc/init.d/networking restart' on the )  Could frequent
re-assignment have caused some sort of problem ??
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091223/41c5d7af/attachment.htm 

From dnsmasq.to.peejay at spamgourmet.com  Sun Dec 27 17:31:46 2009
From: dnsmasq.to.peejay at spamgourmet.com (dnsmasq.to.peejay at spamgourmet.com)
Date: Sun, 27 Dec 2009 12:31:46 -0500 (EST)
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be outside
	that?
Message-ID: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>

dnsmasq.conf has a line:

dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h

(hostmin,hostmax,netmask,leasetime)

If I enable that line, must the dnsmasq server itself be outside that range?

Ie, for the dnsmasq server:
  192.168.0.55 on the LAN interface would be wrong
but
  192.168.0.45 would be ok.

regards
PJ






From ranceh at gmail.com  Sun Dec 27 20:45:58 2009
From: ranceh at gmail.com (Rance Hall)
Date: Sun, 27 Dec 2009 14:45:58 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
Message-ID: <845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>

PJ: see comments below:

On Sun, Dec 27, 2009 at 11:31 AM,  <dnsmasq.to.peejay at spamgourmet.com> wrote:
> dnsmasq.conf has a line:
>
> dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h
>
> (hostmin,hostmax,netmask,leasetime)
>
> If I enable that line, must the dnsmasq server itself be outside that range?
>
> Ie, for the dnsmasq server:
> ?192.168.0.55 on the LAN interface would be wrong
> but
> ?192.168.0.45 would be ok.
>
> regards
> PJ
>

Its usually helpful if your static ip addresses are all in a space
that is allocated to your dhcp pool.

The fact that you don't already know this concerns me greatly.

Dnsmasq is a great product and I've used it successfully and even
tried to contribute when I can.  But getting the most out of dnsmasq
means that you have to have a basic understanding of networking at the
IP level.

Your question indicates that perhaps you don't have that basic understanding.

I fear that you will be creating your own problems over the course of
your dnsmasq installation because you don't really know what you are
doing.

My advice: slow down, do some reading, and try to understand what you
want to do.  If you have a specific question about basic setup, or how
to make dnsmasq fit what you want, then by all means ask.

If I am wrong about my assessment of your question,  I'll apologize.

Rance


From ranceh at gmail.com  Sun Dec 27 23:02:59 2009
From: ranceh at gmail.com (Rance Hall)
Date: Sun, 27 Dec 2009 17:02:59 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
Message-ID: <845c0f80912271502r2ad42493k97c77b877ad74812@mail.gmail.com>

oh crap....


On Sun, Dec 27, 2009 at 2:45 PM, Rance Hall <ranceh at gmail.com> wrote:
> PJ: see comments below:
>
> On Sun, Dec 27, 2009 at 11:31 AM, ?<dnsmasq.to.peejay at spamgourmet.com> wrote:
>> dnsmasq.conf has a line:
>>
>> dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h
>>
>> (hostmin,hostmax,netmask,leasetime)
>>
>> If I enable that line, must the dnsmasq server itself be outside that range?
>>
>> Ie, for the dnsmasq server:
>> ?192.168.0.55 on the LAN interface would be wrong
>> but
>> ?192.168.0.45 would be ok.
>>
>> regards
>> PJ
>>
>
> Its usually helpful if your static ip addresses are all in a space
> that is allocated to your dhcp pool.
>

This statement is WRONG with a CAPITAL WRONG  It should have said:

 Its usually helpful if your static ip addresses are all in a space
that is NOT allocated to your dhcp pool.

Sorry about the confusion.  Can't believe I missed that during my message edit.

The rest of what I said below is still relevant.

> The fact that you don't already know this concerns me greatly.
>
> Dnsmasq is a great product and I've used it successfully and even
> tried to contribute when I can. ?But getting the most out of dnsmasq
> means that you have to have a basic understanding of networking at the
> IP level.
>
> Your question indicates that perhaps you don't have that basic understanding.
>
> I fear that you will be creating your own problems over the course of
> your dnsmasq installation because you don't really know what you are
> doing.
>
> My advice: slow down, do some reading, and try to understand what you
> want to do. ?If you have a specific question about basic setup, or how
> to make dnsmasq fit what you want, then by all means ask.
>
> If I am wrong about my assessment of your question, ?I'll apologize.
>
> Rance
>


From richardvoigt at gmail.com  Sun Dec 27 23:47:11 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 27 Dec 2009 17:47:11 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
Message-ID: <2e59e6970912271547y54f4fbcbj995fdda572fe027b@mail.gmail.com>

On Sun, Dec 27, 2009 at 11:31 AM,  <dnsmasq.to.peejay at spamgourmet.com> wrote:
> dnsmasq.conf has a line:
>
> dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h
>
> (hostmin,hostmax,netmask,leasetime)
>
> If I enable that line, must the dnsmasq server itself be outside that range?
>
> Ie, for the dnsmasq server:
> ?192.168.0.55 on the LAN interface would be wrong
> but
> ?192.168.0.45 would be ok.

In my experience, dnsmasq is smart enough to not give out its own
address, or any addresses mentioned in dhcp-host lines (except to the
computer to which it is reserved).

It may be confusing, because your address pool is smaller than the
dhcp-range suggests, and you'll have fewer leases than expected when
dhcp starts failing, but dnsmasq won't do anything so broken as giving
away its own address.

>
> regards
> PJ
>
>
>
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From ranceh at gmail.com  Mon Dec 28 00:09:23 2009
From: ranceh at gmail.com (Rance Hall)
Date: Sun, 27 Dec 2009 18:09:23 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <2e59e6970912271547y54f4fbcbj995fdda572fe027b@mail.gmail.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<2e59e6970912271547y54f4fbcbj995fdda572fe027b@mail.gmail.com>
Message-ID: <845c0f80912271609w59f220beraa3efce77e5773c@mail.gmail.com>

On Sun, Dec 27, 2009 at 5:47 PM, richardvoigt at gmail.com
<richardvoigt at gmail.com> wrote:
> On Sun, Dec 27, 2009 at 11:31 AM, ?<dnsmasq.to.peejay at spamgourmet.com> wrote:
>> dnsmasq.conf has a line:
>>
>> dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h
>>
>> (hostmin,hostmax,netmask,leasetime)
>>
>> If I enable that line, must the dnsmasq server itself be outside that range?
>>
>> Ie, for the dnsmasq server:
>> ?192.168.0.55 on the LAN interface would be wrong
>> but
>> ?192.168.0.45 would be ok.
>
> In my experience, dnsmasq is smart enough to not give out its own
> address, or any addresses mentioned in dhcp-host lines (except to the
> computer to which it is reserved).
>
> It may be confusing, because your address pool is smaller than the
> dhcp-range suggests, and you'll have fewer leases than expected when
> dhcp starts failing, but dnsmasq won't do anything so broken as giving
> away its own address.
>

Richard:

This is good to know, Ive never even asked dnsmasq to do this, so I
wouldnt have had a clue that it would work correctly.

I always design a static ip pool, a device pool, and a dhcp pool that
way I know whats going on on the network.  and I can easily ignore
devices in searches when Im looking for a problem with an as yet
unidentified pc somewhere.


From dnsmasq.to.peejay at spamgourmet.com  Mon Dec 28 00:47:09 2009
From: dnsmasq.to.peejay at spamgourmet.com (dnsmasq.to.peejay at spamgourmet.com)
Date: Sun, 27 Dec 2009 19:47:09 -0500 (EST)
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
 outside that?
In-Reply-To: <845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
Message-ID: <5249.85.145.170.203.1261961229.squirrel@lavabit.com>

Rance wrote:
> Its usually helpful if your static ip addresses are all in a space
> that is NOT allocated to your dhcp pool.

[Rance's typo is corrected in the above]

I guess I'd better elaborate.

You can set a static ip by not defining it as part of the dhcp server
configuration (dhcpd.conf/dnsmasq.conf), and defining it independently of
the dhcpserver configuraton settings. We can define this as being "not
allocated to your dhcp pool".

The machines that are defined by the dhcpserver configuration are
"allocated to your dhcp pool".

Note that "allocated to your dhcp pool" is inapplicable for the machine
running the dnsmasq server, since it cannot be assigned an ip from the
dhcp service.

My question was about a dnsmasq server that had a static ip possibly in
the lease pool range, If it was in the range, then I expected a conflict
problem. However, dnsmasq is so well-designed that I considered it
possible (though unlikely) that dnsmasq checked if the address was already
in use by the dnsmasq server itself when handing out ips. Simon's answer,
that dnsmasq has no problem with this and knows the ip is used, surprised
me.

The reason I asked the question was because I am scripting network
configurations and I had to cover this case (ie: someone deciding to
define the range of dhcp addresses in such a way that the dnsmasq server
ip address ends up being inside the range). I built a check to prohibit
that, but I wasn't actually sure it was necessary. So I asked the list,
and got that surprising answer from Simon. So I don't need to bother with
the check when using dnsmasq.


On to another issue now: static ips outside the allocated pool.

You seem to be implying that it is better not to define static ips in
dhcpserver config files (that is, it is better to have them outside the
dhcp range pool). Why?

(To be clear, I am thinking here of a lan with a mix of static and dynamic
ips, using a dhcpserver on the lan)

Please explain it, or point me to material that explains why this should
be a best practice.

> The fact that you don't already know this concerns me greatly.

That you are concerned about me not knowing this worries me too.

regards
PJ






From richardvoigt at gmail.com  Mon Dec 28 01:30:51 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 27 Dec 2009 19:30:51 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <845c0f80912271609w59f220beraa3efce77e5773c@mail.gmail.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<2e59e6970912271547y54f4fbcbj995fdda572fe027b@mail.gmail.com>
	<845c0f80912271609w59f220beraa3efce77e5773c@mail.gmail.com>
Message-ID: <2e59e6970912271730r5370d88cof77902b9c050d57a@mail.gmail.com>

On Sun, Dec 27, 2009 at 6:09 PM, Rance Hall <ranceh at gmail.com> wrote:
> On Sun, Dec 27, 2009 at 5:47 PM, richardvoigt at gmail.com
> <richardvoigt at gmail.com> wrote:
>> On Sun, Dec 27, 2009 at 11:31 AM, ?<dnsmasq.to.peejay at spamgourmet.com> wrote:
>>> dnsmasq.conf has a line:
>>>
>>> dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h
>>>
>>> (hostmin,hostmax,netmask,leasetime)
>>>
>>> If I enable that line, must the dnsmasq server itself be outside that range?
>>>
>>> Ie, for the dnsmasq server:
>>> ?192.168.0.55 on the LAN interface would be wrong
>>> but
>>> ?192.168.0.45 would be ok.
>>
>> In my experience, dnsmasq is smart enough to not give out its own
>> address, or any addresses mentioned in dhcp-host lines (except to the
>> computer to which it is reserved).
>>
>> It may be confusing, because your address pool is smaller than the
>> dhcp-range suggests, and you'll have fewer leases than expected when
>> dhcp starts failing, but dnsmasq won't do anything so broken as giving
>> away its own address.
>>
>
> Richard:
>
> This is good to know, Ive never even asked dnsmasq to do this, so I
> wouldnt have had a clue that it would work correctly.
>
> I always design a static ip pool, a device pool, and a dhcp pool that
> way I know whats going on on the network. ?and I can easily ignore
> devices in searches when Im looking for a problem with an as yet
> unidentified pc somewhere.

Well, there's always the case where a customer decides they want a
static address after receiving one from the pool, and they don't want
to change the one they've already got once.  You can do this using
dhcp-host config lines or /etc/ethers -- if that computer requests an
address via dhcp it will always receive the same one, and that address
will not be given to any other computer regardless of whether the
matching host claimed its address via dhcp or had it manually entered.


From perette at barella.org  Mon Dec 28 01:46:35 2009
From: perette at barella.org (Perette Barella)
Date: Sun, 27 Dec 2009 20:46:35 -0500
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <5249.85.145.170.203.1261961229.squirrel@lavabit.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
	<5249.85.145.170.203.1261961229.squirrel@lavabit.com>
Message-ID: <42E116FA-E7A7-4D76-8064-2EA36B5C6336@barella.org>

On 2009?12?27?, at 19:47, dnsmasq.to.peejay at spamgourmet.com wrote:
> On to another issue now: static ips outside the allocated pool.
> 
> You seem to be implying that it is better not to define static ips in
> dhcpserver config files (that is, it is better to have them outside the
> dhcp range pool). Why?

The obvious failure case is that you boot up a device, dnsmasq happens to give out an address via DHCP because it's in the DHCP pool... and at some later time, someone powers on the machine with the same (but statically configured) IP address.  Now you have two machines with the same IP, and suckage will ensue.

So I'd agree with Rance on this one... Keep your static IPs out of the DHCP range.  dnsmasq does what it can to prevent problems, but you're opening the door to headaches.

If, on the other hand, you're talking about pseudo-static IPs where dnsmasq is configured to give out a specific address from within the pool to a particular machine (via dhcp-host)... it would be nice if dnsmasq guaranteed it and wouldn't give the address out to anyone but the specific MAC address, I'd even guess that it's clever enough to behave this way, but I'm not qualified to guarantee it.  Anyone?

Perette



From ranceh at gmail.com  Mon Dec 28 02:44:06 2009
From: ranceh at gmail.com (Rance Hall)
Date: Sun, 27 Dec 2009 20:44:06 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <5249.85.145.170.203.1261961229.squirrel@lavabit.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
	<5249.85.145.170.203.1261961229.squirrel@lavabit.com>
Message-ID: <845c0f80912271844g7655d7bk1bcff948c154888a@mail.gmail.com>

On Sun, Dec 27, 2009 at 6:47 PM,  <dnsmasq.to.peejay at spamgourmet.com> wrote:
> Rance wrote:

>
> On to another issue now: static ips outside the allocated pool.
>
> You seem to be implying that it is better not to define static ips in
> dhcpserver config files (that is, it is better to have them outside the
> dhcp range pool). Why?
>
> (To be clear, I am thinking here of a lan with a mix of static and dynamic
> ips, using a dhcpserver on the lan)
>
> Please explain it, or point me to material that explains why this should
> be a best practice.
>

As far as your scripting goes, good luck.  I'm always interested in
what people do to solve problems like this, and scripts for your own
use are very different from scripts you intend others to use.

Maybe you can share your ideas/problems.  I'd love to hear.

As far as my ip setup goes I always define at least 3 ( and sometimes
more) sub sets of ips in the range my network is going to use.

1)  Servers that need static ip addresses.
2) devices that need static ip addresses.  (like network printers etc.)
3) dhcp pool

If I need to then I can refine this list some more.

company owned dhcp machines and guest dhcp machines

wireless access devices (like access points, etc)

Why do you do this?

Good form
Minimize problems with ip overlap
easier to troubleshoot later

Every network design class I ever attended or taught had a layout like this.

Maybe its as simple as this is what I know.


as far as asking dnsmasq to always assign the same ip address based on
hostname or mac or some other known attribute, that is possible and I
do that all the time.

Temporary test servers are an example of something I allow to get a
dhcp address but use a mac address to make sure it gets a known value.


From richardvoigt at gmail.com  Mon Dec 28 04:50:09 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Sun, 27 Dec 2009 22:50:09 -0600
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server be
	outside that?
In-Reply-To: <42E116FA-E7A7-4D76-8064-2EA36B5C6336@barella.org>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>
	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>
	<5249.85.145.170.203.1261961229.squirrel@lavabit.com>
	<42E116FA-E7A7-4D76-8064-2EA36B5C6336@barella.org>
Message-ID: <2e59e6970912272050r7212ecfej29f5474255d56ab6@mail.gmail.com>

On Sun, Dec 27, 2009 at 7:46 PM, Perette Barella <perette at barella.org> wrote:
> On 2009?12?27?, at 19:47, dnsmasq.to.peejay at spamgourmet.com wrote:
>> On to another issue now: static ips outside the allocated pool.
>>
>> You seem to be implying that it is better not to define static ips in
>> dhcpserver config files (that is, it is better to have them outside the
>> dhcp range pool). Why?
>
> The obvious failure case is that you boot up a device, dnsmasq happens to give out an address via DHCP because it's in the DHCP pool... and at some later time, someone powers on the machine with the same (but statically configured) IP address. ?Now you have two machines with the same IP, and suckage will ensue.
>
> So I'd agree with Rance on this one... Keep your static IPs out of the DHCP range. ?dnsmasq does what it can to prevent problems, but you're opening the door to headaches.
>
> If, on the other hand, you're talking about pseudo-static IPs where dnsmasq is configured to give out a specific address from within the pool to a particular machine (via dhcp-host)... it would be nice if dnsmasq guaranteed it and wouldn't give the address out to anyone but the specific MAC address, I'd even guess that it's clever enough to behave this way, but I'm not qualified to guarantee it. ?Anyone?

It doesn't appear to be part of the contract (e.g. man page) but from
experience I can say that the behavior in every version of dnsmasq
I've used is to reserve addresses mentioned in /etc/ethers or
dhcp-host lines and not give them out to non-matching hosts.


>
> Perette
>
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From simon at thekelleys.org.uk  Mon Dec 28 09:25:34 2009
From: simon at thekelleys.org.uk (Simon Kelley)
Date: Mon, 28 Dec 2009 09:25:34 +0000
Subject: [Dnsmasq-discuss] dhcp-range - must ip address of server
 be	outside that?
In-Reply-To: <2e59e6970912272050r7212ecfej29f5474255d56ab6@mail.gmail.com>
References: <23466.85.145.170.203.1261935106.squirrel@lavabit.com>	<845c0f80912271245l44713fd0ib1462f34f0f018ba@mail.gmail.com>	<5249.85.145.170.203.1261961229.squirrel@lavabit.com>	<42E116FA-E7A7-4D76-8064-2EA36B5C6336@barella.org>
	<2e59e6970912272050r7212ecfej29f5474255d56ab6@mail.gmail.com>
Message-ID: <4B38798E.4030601@thekelleys.org.uk>

richardvoigt at gmail.com wrote:

> It doesn't appear to be part of the contract (e.g. man page) but from
> experience I can say that the behavior in every version of dnsmasq
> I've used is to reserve addresses mentioned in /etc/ethers or
> dhcp-host lines and not give them out to non-matching hosts.

That's correct. Dnsmasq also avoids giving out its own address (pace the 
initial query) and addresses in class C space which end in .0 or .255 
(to work around a windows bug)

Simon.




From pgf at foxharp.boston.ma.us  Wed Dec 30 20:54:42 2009
From: pgf at foxharp.boston.ma.us (Paul Fox)
Date: Wed, 30 Dec 2009 15:54:42 -0500
Subject: [Dnsmasq-discuss] dhcp without dns?
Message-ID: <6292.1262206482@foxharp.boston.ma.us>

i know how to run dns without dhcp, but i have a need for the
opposite.  just in case i've missed something in the man page --
is there a way to run dnsmasq as just a dhcp server?  i'd like to
stick with dnsmasq because i'd rather not have to (re)learn how
to configure the ISC server, or build busybox, or whatever.

(i may be able to work around this by forcing the competing DNS
server on the box to only bind to other interfaces, giving
dnsmasq port 53 access on the interface where i'll be serving
dhcp, but i'd prefer the more general solution.)

paul
=---------------------
 paul fox, pgf at foxharp.boston.ma.us (arlington, ma, where it's 28.8 degrees)


From perette at barella.org  Wed Dec 30 21:05:40 2009
From: perette at barella.org (Perette Barella)
Date: Wed, 30 Dec 2009 16:05:40 -0500
Subject: [Dnsmasq-discuss] dhcp without dns?
In-Reply-To: <6292.1262206482@foxharp.boston.ma.us>
References: <6292.1262206482@foxharp.boston.ma.us>
Message-ID: <7FBA7910-BC03-4FF7-B4FD-6F15FB262219@barella.org>

Quoting the man page:

       -p, --port=<port>
              Listen  on  <port>  instead of the standard DNS port (53).
              Setting this to zero  completely  disables  DNS  function,
              leaving only DHCP and/or TFTP.

Peri

On 2009?12?30?, at 15:54, Paul Fox wrote:

> i know how to run dns without dhcp, but i have a need for the
> opposite.  just in case i've missed something in the man page --
> is there a way to run dnsmasq as just a dhcp server?  i'd like to
> stick with dnsmasq because i'd rather not have to (re)learn how
> to configure the ISC server, or build busybox, or whatever.
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091230/d6ecc067/attachment.htm 

From pgf at foxharp.boston.ma.us  Wed Dec 30 22:05:58 2009
From: pgf at foxharp.boston.ma.us (Paul Fox)
Date: Wed, 30 Dec 2009 17:05:58 -0500
Subject: [Dnsmasq-discuss] dhcp without dns?
In-Reply-To: <7FBA7910-BC03-4FF7-B4FD-6F15FB262219@barella.org>
	(sfid-20091230_164339_433709_5A05C5FF) 
References: <6292.1262206482@foxharp.boston.ma.us>
	<7FBA7910-BC03-4FF7-B4FD-6F15FB262219@barella.org>
	(sfid-20091230_164339_433709_5A05C5FF)
Message-ID: <16923.1262210758@foxharp.boston.ma.us>

perette wrote:
 > Quoting the man page:
 > 
 >        -p, --port=<port>
 >               Listen  on  <port>  instead of the standard DNS port (53).
 >               Setting this to zero  completely  disables  DNS  function,
 >               leaving only DHCP and/or TFTP.

thanks!  don't know how i missed that.

paul


 > 
 > Peri
 > 
 > On 2009???12???30???, at 15:54, Paul Fox wrote:
 > 
 > > i know how to run dns without dhcp, but i have a need for the
 > > opposite.  just in case i've missed something in the man page --
 > > is there a way to run dnsmasq as just a dhcp server?  i'd like to
 > > stick with dnsmasq because i'd rather not have to (re)learn how
 > > to configure the ISC server, or build busybox, or whatever.
 > > 
 > part 2     text/plain                 181
 > _______________________________________________
 > Dnsmasq-discuss mailing list
 > Dnsmasq-discuss at lists.thekelleys.org.uk
 > http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss

=---------------------
 paul fox, pgf at foxharp.boston.ma.us (arlington, ma, where it's 26.8 degrees)


From didster at gmail.com  Wed Dec 30 22:08:14 2009
From: didster at gmail.com (Didster)
Date: Wed, 30 Dec 2009 22:08:14 +0000
Subject: [Dnsmasq-discuss] One lease or another
Message-ID: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>

Hi

Probably an odd request this, but is there anyway of getting dnsmasq
to give a lease to either one machine or another but not both?

So I have a small network of about 7 PCs on a static range.  Each PC
is listed with its MAC address in a dhcp-host directive.  For 2 of the
PCs, I only want one of the 2 machines to be able to get a lease at
any one time.  So if PC1 has a lease, PC2s request will fail and vise
versa.

Is there anyway that can be done?

Thanks
Simon


From ranceh at gmail.com  Thu Dec 31 02:38:10 2009
From: ranceh at gmail.com (Rance Hall)
Date: Wed, 30 Dec 2009 20:38:10 -0600
Subject: [Dnsmasq-discuss] One lease or another
In-Reply-To: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
References: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
Message-ID: <845c0f80912301838w1e57422dj4209e615735c5d7c@mail.gmail.com>

On Wed, Dec 30, 2009 at 4:08 PM, Didster <didster at gmail.com> wrote:
> Hi
>
> Probably an odd request this, but is there anyway of getting dnsmasq
> to give a lease to either one machine or another but not both?
>

I'd agree with you about the "odd" part.

> So I have a small network of about 7 PCs on a static range. ?Each PC
> is listed with its MAC address in a dhcp-host directive. ?For 2 of the
> PCs, I only want one of the 2 machines to be able to get a lease at
> any one time. ?So if PC1 has a lease, PC2s request will fail and vise
> versa.
>
> Is there anyway that can be done?

AFAIK, no, there is nothing that can force this behavior, it might be
against an RFC even.

I'm always curious when people ask questions like this, they are
trying to do something interesting and unique, I'd really like to hear
about why you want to do this.

Rance


From richardvoigt at gmail.com  Thu Dec 31 03:43:09 2009
From: richardvoigt at gmail.com (richardvoigt at gmail.com)
Date: Wed, 30 Dec 2009 21:43:09 -0600
Subject: [Dnsmasq-discuss] One lease or another
In-Reply-To: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
References: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
Message-ID: <2e59e6970912301943i37ab7fa0ta24436d86f5738fb@mail.gmail.com>

On Wed, Dec 30, 2009 at 4:08 PM, Didster <didster at gmail.com> wrote:
> Hi
>
> Probably an odd request this, but is there anyway of getting dnsmasq
> to give a lease to either one machine or another but not both?
>
> So I have a small network of about 7 PCs on a static range. ?Each PC
> is listed with its MAC address in a dhcp-host directive. ?For 2 of the
> PCs, I only want one of the 2 machines to be able to get a lease at
> any one time. ?So if PC1 has a lease, PC2s request will fail and vise
> versa.
>
> Is there anyway that can be done?

Use the dhcp-host lines to assign a tag to these two computers.  Don't
associate an IP address from dhcp-host.
Make a dhcp-range which has a pool of only one address and matches that tag.
Make sure that other dhcp-ranges do not match the tag (as in, #tag or
use the static keyword to not create a pool at all).
Done.

If either PC has a valid lease, the pool is exhausted and the other PC
will be refused.  When the first lease is released or expires, the
pool will have a free address again and the next request will be
accepted.  But please note most clients won't release their lease
during shutdown without special configuration.

>
> Thanks
> Simon
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>


From santiago at zarate.net.ve  Thu Dec 31 23:07:42 2009
From: santiago at zarate.net.ve (Santiago Zarate)
Date: Thu, 31 Dec 2009 18:37:42 -0430
Subject: [Dnsmasq-discuss] One lease or another
In-Reply-To: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
References: <bb09457e0912301408n7200d41ck617b78373d3a8cd5@mail.gmail.com>
Message-ID: <200912311837.52975.santiago@zarate.net.ve>

If i'm not mistaken... and according to the doc... this:

dhcp-host=00:1f:3c:45:ca:c5,00:1e:ec:08:e7:b4,quaoar,192.168.1.140,infinite

should do what you want, so, 2 macs (no idea if you could place more)... 1 ip... 

-- 
Santiago Zarate
santiago at zarate.net.ve
(+58) 4129864175
(+58) 4241073905
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
Url : http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20091231/3e6a461b/attachment.pgp