This probably depends, in part, on how your NFS is configured. An NFS hard mount that is unresponsive could freeze your splunkd -- or at least one or more threads -- until the NFS server does respond again. A soft mount would return an error back to splunk, which may be the more interesting case for your question.