NetApp: Fix race condition in 7-mode iSCSI driver with DFM.
A race condition was discovered in the driver while creating
or deleting multiple volumes in a very short interval of time
(can be reproduced by making successive API calls to Cinder).
This fix decorates the methods responsible for race condition
with a synchronized() decorator. The problem was with the
management software: DFM, and in principle, there should not
be two processes trying to modify the DFM dataset object at
the same time. This patch prevents from such a possibility.
Reviewed: https:/ /review. openstack. org/23077 github. com/openstack/ cinder/ commit/ 3cff1cb24f7627d d913ce02a0f62c7 ef85b5c37c
Committed: http://
Submitter: Jenkins
Branch: master
commit 3cff1cb24f7627d d913ce02a0f62c7 ef85b5c37c
Author: Rushi Agrawal <email address hidden>
Date: Wed Feb 27 06:37:41 2013 -0500
NetApp: Fix race condition in 7-mode iSCSI driver with DFM.
A race condition was discovered in the driver while creating
or deleting multiple volumes in a very short interval of time
(can be reproduced by making successive API calls to Cinder).
This fix decorates the methods responsible for race condition
with a synchronized() decorator. The problem was with the
management software: DFM, and in principle, there should not
be two processes trying to modify the DFM dataset object at
the same time. This patch prevents from such a possibility.
Fixes bug 1091480
Change-Id: I11cfc8868171ac ec356f4140aba52 b8078d109df