HP ProLiant DL380G5-WSS 3.7.0 HP StorageWorks HP Scalable NAS File Serving Sof - Page 196
Clear NLM locks after a power failure on an NFS client, almost immediately.
View all HP ProLiant DL380G5-WSS manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 196 highlights
If necessary, NLM can be enabled; however, you should be aware of the following caveat: File locks granted by the NFS server are cluster-coherent. When a failover occurs, the locks are released by the original server and the client automatically reclaims them on the new server (the backup node). However, during the period after the lock is released, another client or application may compete for and win the lock. Some NFS clients will return an error to the client applications if the lock cannot be reclaimed. Other clients (for example, the Linux 2.6 NFS client) will not return any error. If no error is returned by the client, the application may proceed under the false assumption that the lock has been granted. Data corruption may be the result. To prevent this situation, locking should be enabled only if your clients are partitioned so that all clients needing a particular lock are using the same Virtual NFS Service IP address. If a failover occurs, all of the clients will lose their locks. They can then reclaim those same locks on the new node without conflicts from outside clients. The mxnlmconfig command is used to enable or disable NLM locking in the cluster. The change takes place immediately and may affect clients. The mxnlmconfig command has this syntax: /opt/hpcfs/bin/mxnlmconfig -q|-e|-d|-? The options are as follows: -q Show the current status of NLM locking in the cluster (either enabled or disabled). -e Enable NLM locking in the cluster. No reboot is necessary; the change is effective almost immediately. -d Disable NLM locking in the cluster. No reboot is necessary; the change is effective almost immediately. -? Display a syntax message. Clear NLM locks after a power failure on an NFS client With NLM locking, when NFS clients holding byte-range locks experience a sudden power failure, there is no time for the locks to be cleared as they would be during a normal shutdown. To remedy this, the statd process on the NFS client keeps track of servers whose files contain locks once held by this NFS client. On the NFS client, the directory /var/lib/nfs/statd/sm contains files whose names are IP addresses 196 Configure FS Option for Linux