When using allow_kill="0" on my cluster quorum device, a node that loses access to storage and then regains it gets fenced in RHEL
Issue
- I have
allow_kill="0"set on my quorum device, and if a node loses access to the storage and then regains access, it gets fenced - A node still gets evicted when
allow_killis disabled - With
allow_kill="0", if I disconnect the network cable for my QDisk iSCSI interface, the offending node remains alive until you reconnect the iSCSI cable; once it gets the qdisk again, it reboots (possibly due to an eviction notice written on qdisk by the other node).
Environment
- Red Hat Enterprise Linux (RHEL) 5 or 6 with the High Availability Add On
- Use of a quorum device in the cluster
<quorumd allow_kill="0" ...>set in/etc/cluster/cluster.conf
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
