NetApp senior technical director Mike Eisler, an RFC 5661 editor, believes it could be the middle or end of 2011 before we see any server-software updates available from storage array vendors and a client OS capable of supporting pNFS. "The rationale being that there's no point in having a server-side of a new protocol if you don't have the client side," Eisler said.
So what's next? According to NetApp chief technology officer (CTO) and senior vice president (VP) Brian Pawlowski -- and a co-chair of the Network File System Version 4 Working Group -- the NFSv4.0 and NFSv4.1 codes have the foundation to become a multivendor, heterogeneous and federated global namespace technology.
When a client makes a request for a particular portion of the namespace, and the data the client is seeking isn't in that namespace, the NFSv4.0 and NFSv4.1 Server will direct the client to the optimal namespace to retrieve the data, even if the namespace is part of another vendor's system. Pawlowski said this will allow customers to assemble and deploy a namespace from different vendor systems and deploy it globally.
Larry Jones, VP of marketing at Panasas, sees another benefit from pNFS — file server management. "[pNFS] provides a great platform for server consolidation," Jones said. "Another way to think about it is as scale-out NAS. Instead of having a single NAS Server, what parallel NFS lets you do is consolidate a whole bunch of [NAS] servers. You don't have to go hunting all over the place to find your data. Just one mount point and you get to that data."
Jones believes IT administrators expect a performance boost from pNFS, but the manageability and ease of administration "will surprise them."
This was first published in June 2010