DNS resolution on a managed bridge doesn't work correctly sometimes

Incus
,
1

Incus version: 1:7.0-ubuntu24.04-202605301826

I set up a project with OCI containers. app, database etc. connected to the managed bridge. app connects to the database through the database host. Basically app depends on the DNS resolution on the managed bridge. to be able to connect with the database container. But occasionally the database ⁣resolves to a wrong IP. I was getting errors telling me that the app ⁣tried to connect to a random IP that was not used at all in the Incus’ network. Restarting the database container seemed to fix it, but only once. When the next time it happened it didn’t work. So, I restarted the incus service, and it, again, seemed to work for a while, but the problem reappeared after a short time. Now I restarted the whole machine, and I wonder: if it were to happen again, where should I look to find the reason for it or to be able to provide any useful data for the bug report?

2

You can look at the DHCP leases in /var/lib/incus/networks as that’s what dnsmasq uses to build its DNS zone.

1 Like
3

I’ve had the same troubles with a nested container also have seen dnsmasq errors in sudo journalctl -u incus -f

After a reboot of that container errors went away.

You might wanna have a look at my incus-compose: Incus-compose - the missing docker-compose for incus - 1.0.0-beta11 now available