Comment 15 for bug 1818682

Revision history for this message
Marcus Klein (kleini76) wrote : Re: HAproxy for metadata refuses connection from VM cloud-init

I started once again to debug this issue in our setup. Some more insights came up when looking again into the logs. When a VM is spawned the following two messages appear in the neutron-l3-agent.log:

2019-04-10 14:22:31.503 30752 INFO neutron.agent.l3.ha [-] Router 48d5a991-1327-49c9-88d9-e28b6f48c550 transitioned to backup
019-04-10 14:22:37.492 30752 INFO neutron.agent.l3.ha [-] Router 48d5a991-1327-49c9-88d9-e28b6f48c550 transitioned to master

This is visible through neutron API, too:

Alle 0,1s: neutron l3-agent-list-hosting-router 48d5a991-1327-49c9-88d9-e28b6f48c550 marcus: Wed Apr 10 14:22:35 2019

neutron CLI is deprecated and will be removed in the future. Use openstack CLI instead.
+--------------------------------------+----------+----------------+-------+----------+
| id | host | admin_state_up | alive | ha_state |
+--------------------------------------+----------+----------------+-------+----------+
| 14143f12-1732-4596-8423-42c11d5f73eb | network2 | True | :-) | standby |
| 1ba861cb-b049-47c1-9470-9050975ec5d0 | network1 | True | :-) | standby |
+--------------------------------------+----------+----------------+-------+----------+

The L3 router was previously on network2 and appears then again as active on network2.

My next steps will now be to investigate, why this transition of the L3 router into backup state happens.