Our Environment is consists of X200 6 nodes cluster in OneFS8.0.0.4.

As I got fatal memory errors in the events, a dimm replacement procedure was proceeded on the defective node. Before replacement, I've pulled out the both ext interfaces of the node from an IP address pool for safety. After dimm replacement, the node's interfaces were put back to the IP address pool, and clients got I/O error while 'df -h' or cannot mount back to the node.

the output is similar to:

root@em1:# uname -a

SunOS em1 5.11 11.2 i86pc

root@em1:/# mount -f nfs 110.x.x.127:/ifs/data/data01 /data01

nfs mount: mount: /data01: Stale NFS file handle

root@em3:/# mount -F nfs 110.x.x.126:/ifs/test /test

The problem occuring in other clients from different OS.

Have anyone face this problem in OneFS 8.0.0.4 environment?