[Dnsmasq-discuss] Further thoughts/questions on failover

Chris Green cl at isbd.net
Thu Sep 2 12:45:16 UTC 2021


I'm still ruminating on this problem! :-)

A number of people are using docker containers to manage this, one
live dnsmasq in a container is the default DNS/DHCP for the system and
another dnsmasq container is maintained to keep its files in sync with
the running one.

So far so good.

What I don't follow is what happens if/when failure occurs as I don't
see how the replacement dnsmasq instance can appear at the same
address on the LAN.  So, unless all systems are rebooted they will
lose DNS won't they?

E.g. My default dnsmasq instance runs on a system at 192.168.1.2, if I
had a 'clone' docker container on my desktop machine it would be at
address 192.168.1.3.  So, if 192.168.1.2 dies and I start the dnsmasq
container on my desktop machine it's at 192.168.1.3 but all machines
on the network are configured (until reboot or DHCP reload time) to
use 192.168.1.2.

Am I missing something very obvious here (probably!).  

Just maintaining a Rasberry Pi image and updating the dnsmasq files
on it would actually work better (for me anyway) as I could copy it
out to an actual SD card regularly and all I'd need to do to recover
after a failure would be to plug the SD card into a non-dead Pi and
turn it on.

-- 
Chris Green



More information about the Dnsmasq-discuss mailing list