"pe_fence_node: Node <node> will be fenced because of resource failure(s)" in a RHEL 6 or 7 High Availability cluster with pacemaker
Issue
- A node was fenced when trying to stop a resource
- It looks like
pengine
fenced a node because a resource failed - Why did a node get fenced when a resource failed during a stop operation?
Apr 27 09:44:03 node3 pengine[6594]: warning: unpack_rsc_op: Processing failed op stop for neutron-ha-tool on node1: unknown error (1)
Apr 27 09:44:03 node3 pengine[6594]: warning: pe_fence_node: Node node1 will be fenced because of resource failure(s)
Apr 27 09:44:03 node3 pengine[6594]: warning: stage6: Scheduling Node node1 for STONITH
...
Apr 27 09:44:03 node3 crmd[6595]: notice: te_fence_node: Executing reboot fencing operation (77) on node1 (timeout=60000)
Environment
- Red Hat Enterprise Linux (RHEL) 6, 7, 8 or 9 with the High Availability Add On
pacemaker
- A resource with one or more operations configured with
on-fail=fence
- NOTE:
op stop
is configured withon-fail=fence
by default. If the incident in question happened in response to a failed stop operation, this default setting could be responsible. If it happened on a different operation type (likemonitor
,start
,promote
, etc), then this resource would need to haveon-fail=fence
explicitly configured for that operation type to result in the behavior described in this knowledge solution.
- NOTE:
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.