Comment 24 for bug 1998615

I found this bug report while looking for a solution.

I experience this on a single all-in-one MAAS server running 3.4.0-14321-g.1027c7664

The server has two interfaces. The errors happen twice with the IPv6 of one interface. Two minutes later, they happen again twice with the IPv6 address of the other interface. Then back to the first interface after two minutes.

When they happen, they happen again exactly 30 seconds later with the IP of the same interface. Two minutes later, the other IP.. followed by another disconnect exactly 30 seconds. And so on.

I checked all logs files and there is no errors/warnings before the disconnects happen.

After I restart regiond, the disconnects stop for a while and start again after ~15min in the environment I'm troubleshooting.

After enabling debug logs, I cannot see any errors before the disconnects happen either. There are only a bunch of successful SELECT's to the database and the service monitors checking things. I also don't see any processes dying.