Handle deferred deletion race condition

Bug #2068043 reported by kiran pawar
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Shared File Systems Service (Manila)
In Progress
Undecided
Unassigned

Bug Description

In deferred deletion shares, consider below scenario

00:00: do_deferred_share_deletion runs
00:01: do_deferred_snapshot_deletion runs
00:02 snapshot A of share B gets deleted and put into deferred_deleting
00:03: share B gets deleted and put into deferred_deleting
00:05: do_deferred_share_deletion runs: snapshot A still existing !
00:06: do_deferred_snapshot_deletion runs

We must make sure all snapshot instance are deleted first, before deleting share instance in deferred deletion scenario.
Also, once share will be deferred deleted and error occurred, it should be handled either by manila automatically or by admin only, i.e. end user should never see that share. That means, all errors in deferred_deleting should go to only error_deferred_deleting.

kiran pawar (kpdev)
summary: - Handle deferred deletion snapshot correctly
+ Handle deferred deletion race condition
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to manila (master)

Fix proposed to branch: master
Review: https://review.opendev.org/c/openstack/manila/+/921313

Changed in manila:
status: New → In Progress
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.