[Dnsmasq-discuss] address assignment without checking context?]]

[Simon Kelley]

Benjamin Collar wrote:
> Greetings
> 
> Thank you for the previous information. We're still working on it and
> have come to the following. (The following was forwarded to me by the
> colleague who actually performed the work, so it's written from his
> POV...). Any other tips/leads you could give us would be much
> appreciated! Thanks. Ben.

No problem: see below, I think your colleague has got the network 
configuration slightly wrong.


> 
> Hm... Does not seem to work. Tested with dsnmasq-2.22 on SUSE 9.2:
> eth0 has 10.0.150.10
> eth1 has 10.128.150.10
> 
> ip addr add was issued for each IP also to the other interface. 
> Nevertheless dnsmasq accepts 10.0 address requests only via eth0 and 10.128 requests only via eth1.
>  
> linux:/opt/dnsmasq # ip addr show
> 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
>     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
>     inet6 ::1/128 scope host
>        valid_lft forever preferred_lft forever
> 2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 00:01:02:dd:84:bc brd ff:ff:ff:ff:ff:ff
>     inet 10.0.150.111/15 brd 10.1.255.255 scope global eth0
>     inet 10.128.150.111/32 scope global eth0
                            ^
Problem: the netmask is set to 255.255.255.255, so no addresses apart 
from 10.128.150.111 will be considered to be on the same network. The 
netmask needs to be set the same as on the primary interface.

>     inet6 fe80::201:2ff:fedd:84bc/64 scope link
>        valid_lft forever preferred_lft forever
> 3: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 00:50:da:48:d1:fe brd ff:ff:ff:ff:ff:ff
>     inet 10.128.150.111/15 brd 10.129.255.255 scope global eth1
>     inet 10.0.150.111/32 scope global eth1
                          ^
and again here.
>     inet6 fe80::250:daff:fe48:d1fe/64 scope link
>        valid_lft forever preferred_lft forever


The netmask problem needs fixing before we can say that this approach 
won't work, I think.


HTH

Simon.