LXD Cluster on OVN fails with signal 14 when creating my-ovn

I did my best to set up a system as described in https://linuxcontainers.org/lxd/docs/master/howto/network_ovn_setup/ . I am running ubuntu 22.10 on all tree machines.
I am able to configure the UPLINK network.
When I try to create a the my-ovn network I get the following error:

time lxc network create my-ovn --type ovn
Error: Failed adding router: Failed to run: ovn-nbctl --timeout=10 --db
tcp:,tcp:,tcp: --wait=sb lr-add lxd-net24-lr: signal: alarm clock (2023-04-11T14:54:06Z|00002|fatal_signal|WARN|terminating with signal 14 (Alarm clock))
real 0m11,278s
user 0m0,043s
sys 0m0,050s

What does this error mean in detail?

lxd works fine, I can start containers on all tree cluster members. However with the networking I fail.
I can see the geneve tunnels running between the three nodes. After the failure I can see ports etc on all there nodes …

I checked the ip addresses the OPTS value etc. but could not find an issue.

When I had ovn-central running only one host (and ovn-host on the other two) creating my-ovn worked fine. However only the container running on the host with ovn-central got an ip-address an was able to hit the internet. The other two had the interface but no ip-address.

Can anybody help? How can I analyse this?

Thanks in advance!


Did again a very careful setup and now I can create the my-ovn network.

It may have been a misplaced quote (") in the definition of OVN_CTL_OPTS in /etc/default/ovn-central

Sorry for the confusion.


Good to hear you got it sorted. The OVN commands are run with a timeout to prevent LXD hanging in such situations that the OVN DB is unreachable/unresponsive.