RBD deferred deletion causes undeletable RBD snapshots
Bug #1957804 reported by
Eric Harney
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Cinder |
In Progress
|
Medium
|
Unassigned | ||
Bug Description
When using deferred deletion with the RBD driver, in the following scenario:
1. Create volume A
2. Create snapshot B of volume A
3. Create volume C from snapshot B.
4. Delete volume C. The RBD driver calls trash_move() on the RBD image corresponding to volume C.
5. Attempt to delete snapshot B.
Step 5 will fail because the snapshot is still busy until the trash purge operation occurs.
Calling remove() instead of trash_move() on the volume any time it is possible will help alleviate this issue.
| Changed in cinder: | |
| status: | New → In Progress |
Revision history for this message
| Sofia Enriquez (lsofia-enriquez) wrote | #1 |
| Changed in cinder: | |
| importance: | Undecided → Medium |
| tags: | added: deferred-deletion delete rbd snapshot |
Revision history for this message
| OpenStack Infra (hudson-openstack) wrote Change abandoned on cinder (master) | #2 |
To post a comment you must log in.
Fixed proposed to master https:/