 |
 |
I don't know if John familiar with Panda based analysis. If not, is he interested to run some simple jobs at Panda's ANALY_SLAC site to see if he is comfortable with the procedures? I think David Miller and a few other people know how to do so. Assuming John wants to use Panda and the above works well for him, I think what we can do is to ask the Panda team to set up another site, say, ANALY_SLAC_TEST which has the same parameters as the current ANALY_SLAC, except it will be mapped to a different batch queue at SLAC. This site (ANALY_SLAC_TEST) will not be part of the general US ATLAS analysis operation so that John will not have to compete with other ATLAS grid users. It takes time to setup another site so it would be good to start this process asap. Wei Yang | [log in to unmask] | 650-926-3338(O) > From: "Charles C. Young via RT" <[log in to unmask]> > Reply-To: <[log in to unmask]> > Date: Tue, 10 Mar 2009 16:20:52 -0700 > Cc: <[log in to unmask]>, <[log in to unmask]> > Subject: RE: [SLAC #163429] Request to use few memfs machines for ATLAS > testing > > > Hi Wei, > > A 90% level answer is Panda. The more truthful answer is we are trying to keep > this advancing on all fronts in parallel, so it's not definite. We want to > make a (small) production run. In order to have everything registered and > cataloged, I believe the best way is to go through the official production > mechanism. However, if that does not work for some reason, we would not be shy > about running these jobs "by hand" -- we are desparate for the answers. > Cheers.