Why aren't my Brocade switch ports re-enabled after a fencing event if I am using fence_brocade as my Red Hat Cluster Suite fencing agent?
Issue
- "When I shut down node2, node1 detects its down and disables the brocade switch ports for node2. (which is correct). However when I reboot node2, the brocade switch ports are not enabled automatically, even though node2 is allowed to rejoin the cluster. I need to manually enable them."
- When a cluster is using fence_brocade as a fencing method and a node is fenced, the switch ports that are disabled during the fencing are not re-enabled when the host boots up and they must be manually re-enabled.
- How do I unfence a node after its been fenced by fence_brocade ?
Environment
- Red Hat Enterprise Linux Server 5 (with the High Availability and Resilient Storage Add Ons)
- Red Hat Cluster Suite
- At least 3 nodes in a cluster (this is a requirement for fabric fencing)
There are chances of fence race in a two node cluster without any delay parameter of fencing being unavailable as of now for fence_brocade fence agent. Also, using qdisk adds to the complexity in a two node cluster with fabric fencing. - Fencing method is fence_brocade
- At least 3 nodes in a cluster (this is a requirement for fabric fencing)
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.