I'd like to mention a Grid computing software stack that's available & already reasonably integrated. I'll include a few other comments on HPC software stack adoption for researchers.
It's "Genesis II" http://www.cs.virginia.edu/~vcgr/wiki/index.php/Genesis_II_Installation_Guid...
This has a mixture of licenses since the applications have different sources. I don't know whether any of the included applications are ineligible for FC inclusion based on their license. This is a fairly low-level set of packages for things like job scheduling.
I saw some mention of Condor as a basis for an FC HPC stack. This is not a bad idea at all, and builds on a functional code base.
A standards-based approach is taken by the Genesis II package [within the OGF, which is not an official standards organization like IETF & ISO, but is taking a community standards approach to grid computing & related technologies]. Genesis II incorporates various working & standards-compliant pieces into a partial HPC software stack.
On the theme of 'what will researchers use,' I had some discussion with Jeff about this. Getting researchers to move from their proprietary packages can be challenging. I think a different but highly relevant challenge is the need for an integrated & reasonably well-supported HPC software stack. Researchers would like to get hardware & an HPC stack from one location. Or, at least, get an HPC stack that is fully integrated & supported, not a bunch of independent pieces that they need to stich togeter. That's a nitch I think Nightlife has great potential to address. -- Greg
Dr. Gregory Newby, Chief Scientist of the Arctic Region Supercomputing Center Univ of Alaska Fairbanks-909 Koyukuk Dr-PO Box 756020-Fairbanks-AK 99775-6020 e: newby AT arsc.edu v: 907-450-8663 f: 907-450-8603 w: www.arsc.edu/~newby
nightlife@lists.fedoraproject.org