OpenStack Compute (nova)

compute api delete race condition

Bug #998117 reported by Mark Washenberger
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Compute (nova)
Fix Released
Undecided
Mark Washenberger
OpenStack Compute (nova) 2012.2 "folsom"

Bug Description

If a compute_api.delete() command comes in at about the same time that a compute_manager.run_instance() is happening for that instance, it is possible that the delete() call will think the host has not yet been assigned and simply delete the instance from the database, rather than sending a delete command down to the appropriate manager.

OpenStack Infra (hudson-openstack)
Changed in nova:
assignee: nobody → Mark Washenberger (markwash)
status: New → In Progress
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to nova (master)

Reviewed: https://review.openstack.org/7345
Committed: http://github.com/openstack/nova/commit/91d007426f109dfef2142e28741edd51dcf1fdbc
Submitter: Jenkins
Branch: master

commit 91d007426f109dfef2142e28741edd51dcf1fdbc
Author: Mark Washenberger <email address hidden>
Date: Fri May 11 09:47:10 2012 -0400

    Eliminate a race condition on instance deletes.

    - Add constraint and equality conditions to nova.db[.sqlalchemy].api
    - Use host constraints to ensure the compute api doesn't simply delete
      an instance from the database that a compute manager has already
      started to run.

    This race condition is associated with bug #998117

    Change-Id: Id74192d3e66bea073327342f57ce0f26987efd2d

Changed in nova:
status: In Progress → Fix Committed
Thierry Carrez (ttx)
Changed in nova:
milestone: none → folsom-2
status: Fix Committed → Fix Released
Thierry Carrez (ttx)
Changed in nova:
milestone: folsom-2 → 2012.2
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

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