Generic driver: error deleting shares if service_instance is not created

Bug #1492211 reported by Lucian Petrut on 2015-09-04
10
This bug affects 2 people
Affects Status Importance Assigned to Milestone
Manila
Low
Lucian Petrut

Bug Description

If the share server creation fails when creating a share using the generic driver, the share will get in an error state, unable to be deleted.

Trace: http://paste.openstack.org/show/445275/

In case the server instance_id is missing, the driver should allow the delete operation to continue.

Fix proposed to branch: master
Review: https://review.openstack.org/224102

Changed in manila:
assignee: nobody → Lucian Petrut (petrutlucian94)
status: New → In Progress

Change abandoned by Petrut Lucian (<email address hidden>) on branch: master
Review: https://review.openstack.org/224102

Ben Swartzlander (bswartz) wrote :

Don't know why fix was abandoned....

Changed in manila:
importance: Undecided → Low
Tom Barron (tpb) on 2018-06-19
tags: added: driver
Jason Grosso (jgrosso) wrote :

Lucian are you still working on this defect?

Lucian Petrut (petrutlucian94) wrote :

@Jason there's an abandoned patch that fixes the issue (or at least did, back then). To be honest, although seemingly a trivial fix, it got a -1 that I did not understand and no further reviews in a few months, for which reason I dropped it. If anyone's still interested in it, I can restore it.

Jason Grosso (jgrosso) wrote :

@Lucian let me check what is going on here it looks like Ben said "Don't know why fix was abandoned...."

we could possible still fix this...

I remember this problem from a very long time ago, but the last time I used the generic driver (fairly recently), most "create_share" errors which were due to failure when creating the service_instance were handled very well and the share was deletable just fine.

I'd like to suggest to have more detail in the bug report, such as detailed steps to reproduce on master, as share server creation can fail for so many different reasons that each one needs a different investigation and fix.

Jason Grosso (jgrosso) wrote :

Lucian do we have the info Rodrigo suggested?

Lucian Petrut (petrutlucian94) wrote :

As far as I remember, this was happening when Nova instances failed to start due to insufficient resources (it was a small test environment). Intentionally causing a quota validation might reproduce the issue, assuming it wasn't fixed already.

Jason Grosso (jgrosso) wrote :

Rodrigo is this enough info from Lucian?

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers