Relocating a clustered NFS service to another node causes stale file handles and permission errors on clients when multiple NFS services share an IP in RHEL 5 or 6
Issue
-
When I relocate my cluster service to another node, clients show stale file handles and permission denied errors.
-
On the clients where this NFS filesystem is mounted becomes inactive (stale). Also, when I try to remount it - it will throw an error
mount: failed, reason given by server: Permission denied".
- The node shows mount/umount request failures due to missing exports:
mountd[7301]: refused unmount request from 10.12.15.18 for /nfs1 (/): not exported
Environment
- Red Hat Enterprise Linux (RHEL) 5 or 6 with the High Availability Add On
- Multiple NFS services in /etc/cluster/cluster.conf that share one or more IP resources
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.