Parallel Virtual File System (PVFS)

The Parallel Virtual File System (PVFS) provides high-speed access to file data for parallel applications in an attempt to alleviate the bottlenecks caused by the disparity between I/O performance and processor performance. PVFS provides a cluster-wide consistent name space, enables user-controlled striping of data across disks on different I/O nodes, and allows existing binaries to operate on PVFS files without the need for recompiling.

PVFS stripes file data across multiple disks in different nodes in a cluster. By spreading out file data in this manner, larger files can be created, potential bandwidth is increased, and network bottlenecks are minimized. A 64-bit interface is implemented as well, allowing large (more than 2GB) files to be created and accessed.

In addition to being easy to install, scripts and test applications are included to help with configuration, testing for correct operation, and performance evaluation.

• Compatibility with existing binaries
• Ease of installation
• User-controlled striping of files across nodes
• Multiple interfaces, including a MPI-IO interface via ROMIO
• Utilizes commodity network and storage hardware

PVFS supports the UNIX I/O interface and allows existing UNIX I/O programs to use PVFS files without recompiling. The familiar UNIX file tools (ls, cp, rm, etc.) will all operate on PVFS files and directories as well. This is accomplished via a Linux kernel module, provided as a separate package.

PVFS2 is the latest project from the Parallel Virtual File System development team. As with the original PVFS, PVFS2 is a parallel file system for Linux clusters. It harnesses commodity storage and network technology to provide concurrent access to data that is distributed across a collection of servers. PVFS2 is freely distributed under an LGPL license. PVFS2 is a completely new file system design, sharing almost no code with PVFS 1.x. It is designed to be much more flexible and receptive to new technology while building upon previous research and production experience.

PVFS2 has the following features:

• Ease of installation
• User-controlled striping of files across nodes, and a well defined interface for defining new distribution schemes
• Multiple interfaces, including a MPI-IO interface via ROMIO
• Utilizes commodity network and storage hardware
• Very modular design
• Native support for several popular networking technologies like Myrinet, InfiniBand, and TCP/IP
• Support for user-defined access patterns
• Support for heterogeneous clusters
• Distributed metadata

PVFS1 will be actively maintained and improved until PVFS2 reaches a level of maturity necessary to allow for a smooth transition.

Aspen Systems can install, integrate, and test PVFS on our parallel cluster products. We use only the highest-quality components and advanced software to power your high-performance system.

Contact us to receive a free price quote and to discuss your specific high-performance system requirements. We'll listen to your requests and help you put together a system or cluster with a configuration that works to fit your computing needs.