The past few weeks have been really busy. Of course the flood of data and the prospects of an analysis are driving the work, so it’s a welcome change. I’ll also be heading over to CERN in about a week to go on shift. I’ll be there for about a month, and am excited to get back to where the action is. This is however, if planet earth doesn’t have other plans for me. I have to say, I’ve had flights get cancelled before, but I can honestly say, never because of a volcano eruption.
My sympathies are with the people who are stranded, I know I’ve received about 10 emails saying people had to phone into meetings because they couldn’t get back to CERN. I’ll have to keep everyone update on my status… it looks like most airports should be open on Tuesday (my flight leaves Sunday).
Until then, analyses await… and with those analyses come one word: Grid.
The Grid is the mechanism by which we do large scale computing. Instead of running an analysis on a local computer (which has limited capabilities), I send a “job” to a set of computers which copies the conditions of my work area, and breaks the large job into a bunch of smaller ones so it can run over the data set faster. I say this in confidence… I’ve now sent 3000 jobs to the grid. I <3 analyzing 🙂 I usually send my jobs to BNL (Brookhaven National Lab) and I’ve never had a problem… but now I’m getting the following message:
WARNING : Your job might be delayed since BNL is busy. There are 10398 jobs already queued by other users while 292 jobs are running. Please consider replicating the input dataset to a free site or avoiding the –site/–cloud option so that the brokerage will find a free site
I guess lots of people are running over data. Although it’s annoying to have to wait, it makes me happy to think about all the activity that’s been happening.