Question

A storage administrator is mounting an NFSv3 datastore from a NetApp filer to a group of Linux database servers. The database vendor requires that any temporary loss of connectivity to the NFS server must NOT cause I/O operations to return errors to the application; instead, the client should retry indefinitely until the server responds. Which NFS mount behavior meets this requirement?

Accepted Answer

Mount with the 'hard' option so that RPC calls are retried indefinitely until the server recovers

Answer

Mount with the 'soft' option so that failed RPC calls time out and return an error after retrans attempts

Answer

Mount with the 'intr' option so that I/O is silently discarded during outages

Answer

Mount with the 'nolock' option so that file locking is disabled and I/O never blocks