blazar/bug-1847821-2f9d6e61f438dc4f.yaml at 896ef7f935573dbf297eb47e504487fad3180012 - blazar - OpenDev: Free Software Needs Free Tools

openstack/blazar

Jason Anderson 0346b02cce Fix issue moving hosts back to freepool

If an aggregate fails to be created for some reason, for example because
the requested node is not in the freepool, some cleanup occurs. Part of
the cleanup is ensuring the target node is moved _back_ to the freepool.
This should theoretically "heal" cases where the node wasn't in the
freepool but also wasn't in some lease aggregate.

However, this call to Nova's API refers to the freepool aggregate by its
name, which is not supported here: the ID must be used[1]. This caused
this operation to fail and raise a NotFound, confusingly (because Nova
couldn't find an aggregate with ID='freepool' for example.)

[1]:
https://docs.openstack.org/api-ref/compute/?expanded=add-host-detail#add-host

Closes-Bug: #1847821
Change-Id: I7af4d407b183578617131f0de42becb3dc2bc415

2019-10-18 16:30:11 +02:00

9 lines

339 B

YAML

Raw Blame History

 ---
 fixes:
   - |
     Updates the on_start failure handlers to look up the freepool aggregate by
     ID, not by name, when moving hosts back to the freepool. Fixes issue where
     hosts wind up without any aggregate during lease start failure.  For more
     details, see `bug 1847821
     <https://bugs.launchpad.net/blazar/+bug/1847821>`_.