b77fb72aa9
Both volumes and snapshots have a volume_type_id field in the DB, but when we migrate a volume we leave the snapshots with the old type. This can only happen for retypes that don't do migration, since we cannot do a retype with migration if the volume has snapshots. Leaving the snapshots with the old time makes a mess of the quota usage when we do the retype as well as when we delete the snapshots. This patch fixes the quota issue by making sure the snapshots are retyped as well. This means that we will check quotas for the snapshots when retyping a volume that have them and we will properly reserve and set the quota on retype. Closes-Bug: #1877164 Change-Id: I90e9f85d192e1f2fee4ec8615a5bc95851a90f8e
8 lines
288 B
YAML
8 lines
288 B
YAML
---
|
|
fixes:
|
|
- |
|
|
`Bug #1877164 <https://bugs.launchpad.net/cinder/+bug/1877164>`_: Fix
|
|
retyping volume with snapshots leaves the snapshots with the old type,
|
|
making the quotas wrong inmediately for snapshots, and breaking them even
|
|
more after those snapshots are deleted.
|