In late February, Red Hat announced the release of Enterprise Linux 6.4 with pNFS support. This validates the pNFS standard that Panasas and its founder and chief scientist, Dr. Garth Gibson, have long been championing. While the announcement did not receive much fanfare in the press, Red Hat adoption of pNFS is a key advance for the pNFS standard which until now has lacked a mainstream commercial release vehicle. We believe this to be the most important step forward for pNFS since the protocol was first included in the upstream 3.0 Linux kernel in 2011, after years of effort by engineers from Panasas and other leading storage companies. We salute Red Hat for its strong endorsement of the standard.
Why should you care about pNFS? When used with a high performance scale-out NAS solution like Panasas ActiveStor, pNFS solves many of the problems associated with file system performance and the multiple “islands of storage” encountered with legacy scale-up NAS systems. pNFS enables breakthrough performance by allowing the compute clients to read and write data directly and in parallel to scale-out storage without the need for all I/O being funneled through congested storage server heads. This allows storage systems architected for the pNFS standard to deliver significantly higher scalability, keeping up with the explosive growth in data most companies are experiencing.
At Panasas, we are particularly excited about the opportunity that pNFS represents in building a strong future for our distributed parallel architecture. The Panasas DirectFlow client, a precursor to pNFS for objects, has already demonstrated performance scaling to over 6000 compute clients and 1100 storage nodes in production environments. As momentum continues to build around pNFS, we look forward to working with our customers and prospects as they adopt this industry-changing standard.