The Open Science Grid Conference August 22, 2006Posted by gordonwatts in Uncategorized.
I’m back in Seattle.I flew here last Wednsday. I’ve been here less than a week and I already miss Julia. Paula has been sending me little videos on my phone of Julia’s progress. Today it was her lifting up her neck while on her stomach. This is a big milestone for a baby — getting control of their heads. I return to Ottawa this Sunday.
So, why come back to Seattle? About the same time that we figured out we were going to have a child, I agreed to host the Open Science Grid Consortium meeting here, in Seattle.
The Open Science Grid (OSG) is a loose collection of universities and national labs, all of whom contribute resources to a GRID. The OSG folks provide a software package that each site downloads and uses to configure their site. After that anyone submitting jobs to OSG will can have them run on any number of sites. It isn’t perfect, but it is getting closer to the idea of computing being like the power grid: you submit a job and you don’t care, and the results come back.
One of the big reasons I agreed to have it here is that I would like to see UW more involved in the GRID as a service, rather than the GRID as an intellectual exercise. CHEP, a conference I attended in February, has a huge number of talks concerning the GRID and how to get it to work. This is tough stuff and there are plenty of computer science research careers to be made in getting it right. But there is another side to GRID work: just having access to a large amount of computing on which to do calculations. QCD lattice simulations. LHC Event Reconstruction. Protein Folding. Gene Modeling and Discovery. Those of us that want to use this computing power really don’t care very much about how the work gets done: just that it gets done. UW has some investment in the research aspect, but we are behind the times as far as having access to GRID services for non-computer science researchers. This is what OSG is about: providing a service. UW has started to express some interest, and I have heard (grapevine, big-time) that OSG is one of the GRID’s under evaluation currently. That would be fantastic.
The meeting started today, and goes through Weds. All the talks are slowly being posted online. For someone not part of the OSG this morning was the most interesting as it talked about the science that was being done with the OSG as well as science that is done in a GRID like fashion but not on the OSG (yet).