Why fence_compute stonith resource attempts to fence controller node?
Issue
The fence_compute resource is used to mark nova-compute service on failed compute node as down. However this stonith resource is sometimes invoked to fence controller nodes and this invocation eventually fails (times out).
The timeout is expected in this situation but ideally fence_compute shouldn't attempt to fence controller nodes at all.
Environment
- Red Hat Enterprise Linux 8 (with the High Availability Add On and Resilient Storage Add Ons)
- Pacemaker cluster with fence_compute stonith resource
- Red Hat OpenStack Platform 13
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
