I got this error in the form of a 500 from nova-api while live-migrating tools-redis-6 off cloudvirt1049: https://phabricator.wikimedia.org/P94731 using wmcs-drain-hypervisor
2026-07-06 11:27:08.061 3972 ERROR nova.conductor.manager [None req-56dedffc-25e4-4548-894f-13c8100975d3 novaadmin admin - - default default] Migration of instance 10457456-a444-49e9-9468-ffbdf5a73703 to host None unexpectedly failed.: nova.exception_Remote.GroupAffinityViolation_Remote: Anti-affinity instance group policy was violated
The VM was in ERROR state, I set it to active and tried again, this time it worked:
root@cloudcontrol1006:~# wmcs-openstack server set --state active 10457456-a444-49e9-9468-ffbdf5a73703 Resetting the server state can make it much harder to recover a server from an error state. If the server is in error st atus due to a failed move operation then this is likely not the correct approach to fix the problem. Do you wish to continue? [y/n]:
Host cloudvirt1049.eqiad.wmnet now in maintenance mode. No new VMs will be scheduled in it. ----- OUTPUT for command #1: 'sudo -i bash -c ...r cloudvirt1049'' ----- remember, all this does is set OS_CLOUD=novaadmin wmcs-drain-hypervisor: 2026-07-06 12:13:36,048: INFO: Igoring canary instance canary1049-6 (3e22cbd5-4ad9-4734-a621-741990648b39) wmcs-drain-hypervisor: 2026-07-06 12:13:36,240: INFO: Migrating tools-redis-6 (10457456-a444-49e9-9468-ffbdf5a73703) wmcs-drain-hypervisor: 2026-07-06 12:13:45,501: INFO: current status is ACTIVE; waiting for it to change to ['MIGRATING'] wmcs-drain-hypervisor: 2026-07-06 12:13:46,713: INFO: current status is MIGRATING; waiting for it to change to ['ACTIVE'] wmcs-drain-hypervisor: 2026-07-06 12:14:33,856: INFO: instance 10457456-a444-49e9-9468-ffbdf5a73703 (tools-redis-6) is now on host cloudvirt1080 with status ACTIVE