OpenStack Compute (nova)

  1. Series rocky
  2. Bug #1775934
  3. Comment #10

Comment 10 for bug 1775934

Revision history for this message
melanie witt (melwitt) wrote :

A few months ago, I saw a duplicate of the root cause of this issue downstream [1] where a failure during _set_vm_state_and_notify (example: MessagingTimeout) prevents us from setting the cell_id for an InstanceMapping and destroying the BuildRequest.

The InstanceMapping cell_id part of it was solved (from an MQ failure perspective) by a patch that landed in Ussuri:

https://review.opendev.org/683730

and I think this is the root cause of the problem described in this bug.

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1848737#c1