Back to bug 1021966

Who When What Removed Added
RHEL Program Management 2013-10-22 12:28:53 UTC Target Release --- 4.0
Ayal Baron 2013-10-29 10:25:46 UTC CC abaron
Assignee rhos-maint fpercoco
Eric Harney 2013-11-15 17:30:32 UTC Status NEW ASSIGNED
Depends On 1030659
Assignee fpercoco eharney
Eric Harney 2013-11-22 15:24:46 UTC Status ASSIGNED POST
Eric Harney 2013-11-22 15:26:18 UTC Doc Text Cause: Operations that interact with qcow2 data in the GlusterFS Cinder snapshot driver need to have locking so that more than one operation does not happen concurrently.

Consequence: Corruption of snapshot state -> failed operations.

Fix: Add locking around GlusterFS driver operations in Cinder.

Result: Can issue many snapshot operation requests at once without causing issues.
Eric Harney 2013-11-25 16:52:00 UTC Fixed In Version openstack-cinder-2013.2-3.el6ost
Eric Harney 2013-11-25 17:01:29 UTC Status POST MODIFIED
errata-xmlrpc 2013-11-26 19:22:56 UTC Status MODIFIED ON_QA
Haim 2013-12-02 10:04:22 UTC QA Contact hateya tshefi
Don Domingo 2013-12-11 01:31:25 UTC CC ddomingo
Doc Text Cause: Operations that interact with qcow2 data in the GlusterFS Cinder snapshot driver need to have locking so that more than one operation does not happen concurrently.

Consequence: Corruption of snapshot state -> failed operations.

Fix: Add locking around GlusterFS driver operations in Cinder.

Result: Can issue many snapshot operation requests at once without causing issues.
GlusterFS driver operations in Cinder now use locking when performing multiple snapshot operations. Locking ensures that multiple operations interacting with qcow2 data in the GlusterFS Cinder snapshot driver do not happen concurrently.

In previous releases, the lack of locking sometimes resulted in failed operations or even corruption of the snapshot state whenever multiple snapshot operations were issued.
Don Domingo 2013-12-11 23:38:57 UTC Doc Text GlusterFS driver operations in Cinder now use locking when performing multiple snapshot operations. Locking ensures that multiple operations interacting with qcow2 data in the GlusterFS Cinder snapshot driver do not happen concurrently.

In previous releases, the lack of locking sometimes resulted in failed operations or even corruption of the snapshot state whenever multiple snapshot operations were issued.
GlusterFS driver operations in the Block Storage service now use locking when performing multiple snapshot operations. Locking ensures that multiple operations interacting with qcow2 data in the Block Storage service's GlusterFS snapshot driver do not happen concurrently.

In previous releases, the lack of locking sometimes resulted in failed operations or even corruption of the snapshot state whenever multiple snapshot operations were issued.
Dafna Ron 2013-12-12 12:56:09 UTC Status ON_QA VERIFIED
Scott Lewis 2013-12-16 18:40:48 UTC Target Milestone --- rc
errata-xmlrpc 2013-12-19 17:36:18 UTC Status VERIFIED RELEASE_PENDING
errata-xmlrpc 2013-12-20 00:30:40 UTC Status RELEASE_PENDING CLOSED
Resolution --- ERRATA
Last Closed 2013-12-19 19:30:40 UTC
Eric Harney 2013-12-23 17:47:56 UTC See Also https://bugzilla.redhat.com/show_bug.cgi?id=1045428
Eric Harney 2013-12-23 18:01:50 UTC See Also https://bugzilla.redhat.com/show_bug.cgi?id=1045428
Link ID Red Hat Bugzilla 1021966
Perry Myers 2016-04-26 15:19:31 UTC CC pmyers

Back to bug 1021966