Back to bug 1033123

Who When What Removed Added
Allon Mureinik 2013-11-21 15:33:46 UTC Depends On 1009812
Link ID oVirt gerrit 21386
Link ID Red Hat Knowledge Base (Solution) 540033
Link ID oVirt gerrit 21500
Link ID oVirt gerrit 19871
Link ID oVirt gerrit 21501
Link ID oVirt gerrit 21291
Status NEW ASSIGNED
errata-xmlrpc 2013-11-26 10:34:48 UTC Link ID RHEV gerrit 10382
Status ASSIGNED POST
Link ID RHEV gerrit 10381
Status POST MODIFIED
Status MODIFIED ON_QA
Aharon Canan 2013-11-28 10:19:01 UTC Status ON_QA VERIFIED
Charlie 2013-11-29 04:23:20 UTC CC cboyle
Doc Text When a hypervisor was rebooted all LVs which were part of a FC storage domain were automatically activated. This was cause some issues on occasion, as logical volumes should by activated only on request of the engine if needed (VM start), and deactivate immediately when the LVs are not needed (VM stop). This happened because when using FC storage, physical volumes are connected during boot, and vdsm. Logical volumes are auto-activated by both /etc/rc.sysinit and /etc/init.d/netfs startup scripts. These logical volumes did not pick changes done by the SPM on the storage, which could eventually lead to data corruption when a vm is trying to write to the logical volume with stale meta data.

The fix checks all vdsm logical volumes during lvm bootstrap and deactivates them if possible. Special logical volumes are refreshed, since they are accessed early when connecting to storage pool, possibly before lvm bootstrap is done. Open logical volumes are skipped because it is assumed that they use correct meta data when opened.
errata-xmlrpc 2013-12-18 08:02:06 UTC Status VERIFIED RELEASE_PENDING
errata-xmlrpc 2013-12-18 13:58:49 UTC Status RELEASE_PENDING CLOSED
Resolution --- ERRATA
Last Closed 2013-12-18 08:58:49 UTC
Allon Mureinik 2016-02-10 17:36:47 UTC oVirt Team --- Storage

Back to bug 1033123