Tom Parrott

I'm thinking perhaps it would be useful to add the instance's DNS record to the OVN DB at `Add()` time (if static IPs are assigned) as this would give us...

As the connection appears to not being established, I'd be tempted to think its the connection timeout triggering. Can you show the output of `sudo ss -tlpn | grep lxd`...

The problem is here ``` t=2022-02-03T00:19:08+0100 lvl=dbug msg="Database error: Failed updating heartbeat time for member \"10.0.1.30:8443\": database is locked" t=2022-02-03T00:19:08+0100 lvl=dbug msg="Retry failed db interaction (Failed updating heartbeat time for...

Also do you still see the heartbeat errors with latency introduced when not trying to move the instance?

Is it possible you are saturating your link preventing the DB operations from progressing in a timely manner? Although I suppose the transfer hasn't started yet.

Thanks for this info. I've reproduced this exact issue, investigating now...

So looking into this, I believe this is simply caused by time it takes for the target member to perform the necessary DB queries across the network to the member...

Removing the bug label, as its not clear at this time whether we are supporting this type of latency. The reason 4.20 started seeing this issue is that it added...

There are likely lots of these sorts of operations that we should optimise to provide reasonable utility over a high latency link (most of them won't have timeouts, but will...

Ive removed a few more unnecessary db queries recently as part of moving storage db record management into the storage package so will revalidate if that's improved things.