The Leading Source for Global News and Information Covering the Ecosystem of High Productivity Computing
May 25, 2007
Panasas Gives Parallel NFS a Boost
This week HPC storage vendor Panasas announced that they are making their parallel file system client software, called DirectFLOW, publicly available via open source. Their goal is to accelerate the adoption of the Parallel Network File System (pNFS) standard, which is to be released as an extension to the Network File System version 4 protocol, and will be technically known as NSF version 4.1.
HPCwire spoke with Larry Jones, Vice President of Marketing for Panasas, to get the details of the announcement.
According to Jones, pNFS is the first performance upgrade to NFS in two decades, since the protocol was developed by Sun Microsystems. Jones explained that with NFS, you have a tier of servers between the clients and the storage systems and all of your data is read in or read out through those servers. With pNFS, the servers get pulled off to the side; they still control access to the storage systems to make sure that the data is properly protected, but the clients get both direct and parellel access to that storage system -- that's what accounts for performance boost.
There exists a huge imbalance between the compute side and the I/O architectures that pNFS resolves as it allows for parallel and direct access from the client to the storage device.
Jones explained that "NFS with its serial one-lane-in, one-late-out architecture doesn't meet the needs of the HPC community and increasingly the enterprise market" and the solution is pNFS.
Increased use of parallelism from cluster and multi-core processor deployments is driving the need for parallel storage to improve application performance. pNFS offers increased scalability, performance and manageability. In addition, users will have the comfort of not being locked in a by a single vendor.
Jones explained that the increased manageability is especially a boon for HPC customers. pNFS allows a single global namespace instead of having a mount table with 25 NFS servers on it. A single mount point allows for very scalable storage system, making it easier to maintain performance without all the load-balancing issues.
Although the same advantages are currently available in proprietary parallel systems like IBM's GPFS, IBRIX's Fusion Parallel File System, and Panasas' own PanFS, these systems are incompatible with each other. pNFS offers a single standard that should encourage customer deployments of parallel storage solutions.
Panasas has taken a three-step strategy to promote adoption of parallel storage: to create great products that show the potential for parallel storage, work those into standards so the industry can evolve, and, for the end users to benefit, work with the ISV community so that they can develop their software accordingly.
Panasas believes that in order for there to be a big market for parallel storage there needs to be a standard. They hope the company's efforts will energize the market. Jones said that Panasas would "like to see the whole market evolve, so
that we can move from that single-lane NFS to parallel NFS and do it
across the entire industry."
To help accomplish this goal, Panasas has fostered dedicated development resources. They have recently opened an R&D center in TelAviv, Israel, where they have engineers specifically working on the Linux implementation of pNFS, i.e., inserting pNFS into the Linux kernel. There is also a group at the University of Michigan, called The Center for Information Technology Integration (CITI), working towards the same goal. Jones expressed confidence that the pNSF standard will be included in future Linux distributions, perhaps by next year sometime.
Panasas' founder and CTO, Dr. Garth Gibson, a couple of their engineers, and Peter Corbett of Network Appliance wrote the first draft of the pNFS spec back in 2004. They based the architecture on Panasas' DirectFlow protocol. Since then, Panasas has been evolving their software in concert with the standard. Other vendors working on the pNFS standard include Network Appliance, IBM, EMC and Sun.
The Internet Engineering Task Force (IETF) NFSv4 subcommittee is reviewing pNFS as part of NFS version 4.1 and is expected to be finished in September. As soon as that happens, Panasas customers will be able to upgrade their software to the standard. Support is expected on Linux, Windows, and leading Unix versions such as AIX and Solaris. Ultimately, Jones hopes that users will be able to get the technology from their Linux distribution (SUSE, Red Hat, etc.).
Jones concluded by saying: "We think that parallel is the wave of the future. NFS hasn't had a performance kicker in twenty years and it's about time it got one. We believe that parallel is the way to go and we've been leaders in that."
Article Tools
Share Options
(Digg, Technorati, more)
Subscribe
Sponsored Links
RSS Feeds
Feeds by Topic
- Applications
- Developer Tools
- Interconnects
- Middleware
- Networks
- Processors
- Storage
- Systems
- Visualization
Feeds by Industry
Feeds by Content Type
Top Headlines
3D Seismic Data: Taking a Smarter Approach to Interpretation
Jul 09 | Engineer Live | The demand for computational tools to underpin the 3D seismic interpretation process has never been more apparent. Read more...
Engineering Unemployment Soared in 2Q to 8.6%
Jul 08 | EE Times | Unemployment for U.S. engineers has reached record levels, according to government figures. Read more...
Gartner Adjusts 2009 IT Spend Downward Again
Jul 08 | Network World | Global spending for 2009 projected to drop 6 percent, for a total of $3.2 trillion. Read more...
Concurrent and Parallel Are Not The Same
Jul 08 | Linux Magazine | Portability or efficiency? Neither is guaranteed when writing explicit parallel code. Read more...
800 TFLOP Real-Time Ray Tracing GPU Unveiled, Not for Gamers
Jul 07 | Ars Technica | Japanese company builds custom ASIC to accelerate real-time ray traced rendering for the auto industry. Read more...
Featured Whitepapers
Building High Performance Computing in a Green and Modular Solution Building Block
Apr 14 | | Many HPC IT departments are feeling the rising pressure to deliver more capacity computing and performance while trying to reduce the total cost of ownership. This white paper discusses how an environmentally-friendly and open-standards HPC building block based computing system using flexible interconnect options helps address capacity computing needs.
Multimedia
Webcast: Dell Expands HPC Access and Adoption with Intel Cluster Ready Program
Source: Addison Snell, GM/VP, Tabor Research; sponsored by Dell
Many organizations that could benefit from the use of HPC clusters find that it is complicated to get the systems up and running because of limited IT resources or the complexities of the clusters themselves. Learn how the Intel Cluster Ready program, for which Dell was an original partner, seeks to address this challenge for entry level and mid-range HPC users.
Video White Paper: Architecting a Better Network Storage Solution
BlueArc's Titan architecture represents an evolutionary step in file servers by creating a hardware-based file system that can scale bandwidth, IOPS, and overall data capacity well beyond conventional software-based devices. With its ability to virtualize a massive storage pool of up to four usable petabytes of tiered storage, Titan can scale with growing data requirements, offering a competitive advantage for businesses, researchers, or other enterprises seeking to better manage data growth while still ensuring optimal performance.
Webcast: HPC Development Solutions: Sun Studio & Sun HPC ClusterTools
Sun Studio Compilers and Tools and Sun HPC ClusterTools allow you to create high performance parallel applications for OpenSolaris, Solaris and Linux. Sun Studio Express 11/08 includes MPI performance analysis capabilities and full OpenMP 3.0 compiler support. Learn about all this and the latest in Sun HPC ClusterTools 8.1.
Newsletters
Stay informed! Subscribe to HPCwire email Newsletters.
HPC Job Bank
-
HPC Systems Administrator
McGill University -
High Performance Computing Data Center Manager
National Renewable Energy Laboratory -
High Performance Computing Systems Administrator
Idaho National Laboratory -
Solution Engineer- HPC
Dell Inc -
Parallel Programmer
Cardiff University -
Principal Software Research Engineer
Indiana University -
Post Doctoral Researcher - Exascale Stream Computing
IBM - Visit the HPCwire Job Bank
Featured Events
-
July 13-16, 2009
WORLDCOMP'09
Las Vegas , NV
USA
-
August 24-28, 2009
Data Mining: Levels I, II, & III
Denver , CO
USA