[Dnsmasq-discuss] Option 12 hostname sent to RPi seems incorrect

[Geert Stappers]

On Wed, Oct 20, 2021 at 07:50:58AM +0530, Shrenik Bhura wrote:
> On Tue, 19 Oct 2021 at 21:01, Dominik Derigs <dl6er at dl6er.de> wrote:
> 
> > Hey Shrenik,
> >
> > On Tue, 2021-10-19 at 17:20 +0530, Shrenik Bhura wrote:
> > > Option 12 hostname sent to RPi is the first octet of the IP
> > > address. This seems incorrect.
> >
> > This can very well be desired behavior (if the hostname is set to
> > "192"). Can you quote your configuration and related files (e.g.
> > if you are using dhcp-hostsfile or dhcp-hostsdir)?
> >
> > Best,
> > Dominik
> >
> 
> Hi Dominik,
 
Hello Archive visitor,
Hello Mailinglist subscriber,


> There isn't anything specific in the configuration that serves the
> hostname. I am aware of this feature and when defined it works as desired.
> But when nothing is explicitly defined to serve hostnames, it is defaulting
> to 192 as visible in the logs, and only in the case of an RPi.
> When checked on the client with the hostname command, it shows the full
> hostname - 192.168.67.53

Such hostname do I call  IPaddress.


> The same configuration works fine with a x86 system as no hostname is being
> served when none is defined.
> 
> Also it would be ideal to improve the logs as well to reflect the full
> hostname, if at all.
> 
> Have attached the config files involved.

Make a packet capture and put the .pcap on a server.
Then tell here how to fetch the file.
Next can those who care examine the .pcap with their tool.
(wireshark / tcpdump)
 
> Once again, do note that this is the case only for an RPi. So there is some
> specific logic in dnsmasq coming into play for an RPi that seems to be
> serving the IP address as the default hostname when none is present.

So beside the
   /sbin/tcpdump  -w rpi.pcap ether host dc:a6:32:78:98:ae
also
   /sbin/tcpdump  -w worksfine.pcap ether host MAC_address_of_known_good
 

> Best,
> Shrenik

Please make it possible to read in the discussion.  Show the world
that you understand the concepts like "timezone" and "one to many
communication".  In other words:  the response goes below previous text.


Groeten
Geert Stappers
-- 
Silence is hard to parse