Hi Toni,

Yes, that is one thing to try which is probably how the issue happened, since this error leg is very infrequent.

The other thing I was thinking is we might be able to manually trigger this code path by removing (or moving) the neutron-keepalived-state-change script out of the way and create an HA router.  I think that would trigger an exception that would delete the router and go through this path.  You'd just have to watch for "Error while initializing router $id" in the l3-agent log.

-Brian

I don't think this traceback is related to the change, it actually looks like this bug:

https://bugzilla.redhat.com/show_bug.cgi?id=1470784

https://bugs.launchpad.net/neutron/+bug/1706750

The code in OSP10 isn't exactly the same, so I'm not sure it can just be picked there, we'd have to look further.

I would have to look into that bug separately as I'm not sure that backport won't cause other problems.

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2018:0357