Calsoft Whitepaper - Parallel NFSs
A number of distributed file systems appeared in 1980s in the Unix community. NFS, RFS, AFS were a few of those. By far NFS has been the most successful, being used on thousands of Unix and non-Unix operating systems throughout the world. pNFS is a significant extension of the NFS, being a parallel implementation of the NFS that parallelizes file services. In the context of growing computing power that enables a shift away from the traditional serial organization of the computer algorithms, number of attempts are focusing on parallelization of work within and without of a system. Garth Gibson and Peter Corbett published a problem statement in July 2004 that went on to be the starting point for evolution of pNFS. Rapidly changing hardware configurations, deployment and a desire to make optimal usage of available processing power highlighted the problem of limited bandwidth to NFS servers. 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. While NFSv4 allows file system migration, it does not provide a mechanism that supports multiple servers simultaneously exporting a single writable file system.
- Related works
- System overview
- Implementation details