Parallel NFS – A Scalable, High Performance NFSv4 Server using Parallel-IO

A number of distributed file systems appeared in 1980s in the Unix community including NFS, RFS, and AFS. By far, NFS has been the most successful, being used on thousands of Unix and non-Unix operating systems. This paper discusses pNFS which is an implementation of NFS that parallelizes file services.

Gibson and Corbett published a problem statement in July 2004 that went on to be the starting point for the evolution of pNFS. Rapidly changing hardware configuration, deployment, and a desire to make optimal use of available processing power highlighted the problem of limited bandwidth. The bandwidth limitation exists because an NFS server has limited resources like network, CPU, memory, and disk I/O. However by design, access to any one file system through the NFSv4 protocol requires that a single server be accessed.

The solution outlined in the paper proposes the parallelization of file services, by enhancing NFSv4 in a minor version.

To download this whitepaper, click the button below