cinder-volume process is incorrectly active on two nodes after failer of docker service
Issue
- cinder-volume process is incorrectly active on two nodes after failer of docker service.
- I tried to check pacemaker behavior when docker service suddenly disappeared by 'systemctl stop docker'. pacemaker relocates the cinder-volume service to other node. However, the node whose docker service is stopped still runs cinder-volume. Finally, cinder-volume service runs on 2 nodes.
Pacemaker didn't trigger the fence to the node.
Environment
- Red Hat OpenStack Platform 13
- resource-agent
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.