lvm resource fails to start with "<node> owns <vg> and is still a cluster member" after a reboot in RHEL 6
Issue
- If I freeze my services and reboot all nodes, lvm resources fail to start back up with "nodeX owns VG and is still a cluster member"
- During a scheduled maintenance two node cluster was rebooted, but the cluster services failed to start automatically and following error messages were observed:
Dec 19 02:57:19 node1 rgmanager[8944]: [lvm] node2 owns myVG and is still a cluster member
Dec 19 02:57:25 node1 rgmanager[4367]: Services Initialized
Dec 19 02:57:26 node1 rgmanager[9285]: [lvm] Starting volume group, myVG
Dec 19 02:57:26 node1 rgmanager[9322]: [lvm] node2 owns myVG and is still a cluster member
Dec 19 02:57:26 node1 rgmanager[9344]: [lvm] Someone else owns this volume group
Dec 19 02:57:26 node1 rgmanager[4367]: start on lvm "myVG" returned 1 (generic error)
Dec 19 02:57:26 node1 rgmanager[4367]: #68: Failed to start service:myService; return value: 1
Environment
- Red Hat Enterprise Linux (RHEL) 6 with the High Availability Add On
rgmanager- HA-LVM with tagging
locking_type=1in/etc/lvm/lvm.conf<lvm/>resource in/etc/cluster/cluster.conflv_nameblank or unspecified in<lvm/>resource
- Services frozen with
clusvcadm -Zbefore reboot or shutdown of rgmanager, or nodes were shut down "uncleanly" without properly stopping service containing<lvm/>resource
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.