Improve filter scheduler capacity update information

Bug #1180976 reported by John Griffith
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
Cinder
Fix Released
Low
John Griffith

Bug Description

Recent improvement/speed-ups to Tempest test exposed a race-condition in volume delete/create sequences where the capabilities update interval of 60 seconds is not sufficient.

We should look at adjusting the polling interval for the gate tests for this and also possibly consider doing a forced update when we encounter no hosts available from the filter scheduler and retry just to avoid any false errors here.

Changed in cinder:
importance: Undecided → Low
Revision history for this message
Giulio Fidente (gfidente) wrote :

A different approach I would suggest is: given that free_capacity_gb seems to be downsized every time a new volume is allocated (before the volume update interval comes into play), it should also be upsized every time a volume is deleted.

Revision history for this message
John Griffith (john-griffith) wrote :

Indeed, that's probably the correct approach

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to cinder (master)

Fix proposed to branch: master
Review: https://review.openstack.org/29599

Changed in cinder:
assignee: nobody → John Griffith (john-griffith)
status: New → In Progress
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to cinder (master)

Reviewed: https://review.openstack.org/29599
Committed: http://github.com/openstack/cinder/commit/642526013c855e88873ecb96e8680698e095d83a
Submitter: Jenkins
Branch: master

commit 642526013c855e88873ecb96e8680698e095d83a
Author: John Griffith <email address hidden>
Date: Fri May 17 14:03:38 2013 -0600

    Update/Publish volume service updates on delete.

    The Filter scheduler weighing function was failing because
    we were only updating the capacity info either on the periodic
    60 second interval or after succesful creates.

    The result was that if you exceeded your capacity within the
    update interval you would no longer be able to create volumes even
    if you deleted existing volumes until the update was performed and
    published.

    Since we're updating on on create, there's no reason why we shouldn't
    do the same on delete for the applicable volume-service.

    The periodic updates are still important for other purposes so they're left
    alone, but with this change we keep things up to date so the filter scheduler
    has the correct picture of the world.

    Fixes bug: 1180976

    Change-Id: Ic551f512c5a0cf51bafa8b60f0b6b7debbe22692

Changed in cinder:
status: In Progress → Fix Committed
Thierry Carrez (ttx)
Changed in cinder:
milestone: none → havana-1
status: Fix Committed → Fix Released
Thierry Carrez (ttx)
Changed in cinder:
milestone: havana-1 → 2013.2
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.