Clustered lvm resource failed with "[lvm] WARNING: <vg> should not be active" in RHEL 6
Issue
- A status check on an
lvm
resource randomly failed with "should not be active" - Service Group automatically failed over on alternate node.
Dec 17 01:09:41 node1 rgmanager[56937]: [lvm] WARNING: myVG should not be active
Dec 17 01:09:41 node1 rgmanager[57048]: [lvm] WARNING: node1.example.com does not own myVG
Dec 17 01:09:41 node1 rgmanager[57139]: [lvm] WARNING: Attempting shutdown of myVG
Dec 17 01:09:43 node1 rgmanager[56210]: status on lvm "myVG" returned 1 (generic error)
Dec 17 01:09:43 node1 rgmanager[56210]: Stopping service service:myServices
Dec 17 01:12:36 node1 rgmanager[27559]: [lvm] Stripping tag, node1.example.com
Environment
- Red Hat Enterprise Linux (RHEL) 6 with the High Availability Add On
rgmanager
HA-LVM
with tagginglocking_type = 1
in/etc/lvm/lvm.conf
- Clustered flag is not set on VG
lv_name
is blank or not specified in<lvm>
resource
- NOTE: A similar issue exists in certain releases of
resource-agents
that can cause this to happen if a node is rejoining. Please determine if that solution is applicable before following any recommendations here.
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.