As part of the development team, we’re announcing PVFS2 version 1.0 here in Pittsburgh at the SC2004 conference!PVFS2 is a GPL/LGPL based parallel file system for cluster-based applications. It logically groups any number of storage servers into a coherent file system for use by client nodes, specifically tailored to handle efficient access to large shared files. PVFS2 supports access via an MPI-IO interface for high-performance parallel applications, but you can still mount it like a regular GNU/Linux file system for traditional serial applications and managment. The PVFS2 project is conducted jointly between the Parallel Architecture Research Laboratory at Clemson University and the Mathematics and Computer Science Division at Argonne National Laboratory. Please feel free to give it a try!
How does this compare to Intermezzo?
…to OpenBFS ?
OpenBFS isn’t a parallel filesystem, so it’s an apples to kumquats comparison.
speaking of OpenBFS, what ever happened to that? Everyone was saying it was quick and superbly wonderful 3 years back!
So is this PVFS2 really only applicable or useful to high end machines and clusters etc? Just wondering if it will have an impact on the Desktop machine.
Sounds great though.
OpenBFS is still the native FS for Haiku, and SkyOS adopted it with some changes not that long ago under the name SkyFS.
What ever happened to that filesystem called Tux from a few years back, it was all the rage at one point…but it seems to have dropped off the face of the earth?!
mv * > /dev/null
This FS is under LGPL/GPL, is it part of it under one license and part of it under another or can you choose which license or what…?
What license is OpenBFS under? When will we be seeing OpenBFS in common Linux distros?
> What license is OpenBFS under?
The MIT license. More info over at http://www.bug-br.org.br/openbfs/
> When will we be seeing OpenBFS in common Linux distros?
Not for a long time, i guess. If I recall correctly, the OpenBFS needs the kernel to be aware of its capabilities.
What is PVFS2?
PVFS2 is an open-source, scalable parallel file system targeted at production parallel computation environments. It is designed specifically to scale to very large numbers of clients and servers. The architecture is very modular, allowing for easy inclusion of new hardware support and new algorithms. This makes PVFS2 a perfect research testbed as well.
http://www.pvfs.org/pvfs2/pvfs2-faq.html
What is InterMezzo?
InterMezzo is a new distributed file system with a focus on high availability. InterMezzo will be suitable for replication of servers, mobile computing, managing system software on large clusters, and for maintenance of high availability clusters.
http://www.inter-mezzo.org/
The difference appears to be the difference between a parallel computational environment with an associated filesystem to assist with the parallel Input/Output vs. a distributed network.
While the parallel filesystem is writing data to one file (for example) that is proportionately split across multiple nodes (hard drives / processors / computers / machines), the distributed network is copying that one whole file on multiple servers for reliability purposes.
However, I do not know about Intermezzo’s claim concerning “managing system software on large clusters”.
Parallel processing is used when the computation is larger than one processor can efficiently handle. When multiple processors are used to process the same computational problem, these multiple processors will often require either the same data set or different parts of the same data set. The parallel Input/Output allows the multiple processors to access (read/write) the data in an efficient parallel manner by splitting the file into one piece for each node. This results in having one file, with one filename, spread across multiple nodes. Each node with its own data subset within that one, proportionally split apart, file.
Sequential Input/Output (one file on one hard drive) is the traditional method but possesses too large of a bottleneck for efficient parallel computation. Having a specialized filesystem will assist with removing many of the low level file access code to simplify the code writing for the computations.
This is in regards to my post about the Tux filesystem from a few years ago, the name was not Tux but actually TuxFS; this filesystem was supposed to have some amazing features…but as I said before it’s dropped off the face of the earth.
mv * > /dev/null
It was actually called Tux2.
http://www.uwsg.iu.edu/hypermail/linux/kernel/0208.3/0336.html
That’s what I meant, Tux2, not TuxFS…although TuxFS sounds like a better name in my opinion!
mv * > /dev/null