[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Condor-users] Optimizing Simulation
- Date: Fri, 01 May 2009 14:54:54 +0200
- From: Max Raith <max_raith123@xxxxxx>
- Subject: [Condor-users] Optimizing Simulation
Hey Condor Users,
due to your help I was able to build up a Condor Pool with a few windows machines. I can run simple test jobs and it works fine. Well. my task for now is, that I have a huge amount of data files(>1000). Each of these filles is used as input for a simulation programm. The Simulation itself only needs a few seconds to execute. The Simulation Programm produces a file where the results are stored. As a second step the simulation results have to be analysed by a python script(needs another few seconds). Again, this python script produces an output file, which is the import one. These two steps, simulation and analyse should run automatically.
I thougt about creating a .sub file for each Job1 (Job1 = 1 Datafile ->Simulation.exe->1 Resultfile) and Job 2(Job2 = 1 Resultfile-> python.py->final Resultfile)
and linking both via dagman. Is this this a good choice? Or do you see a better way of doing it?
Is it possible, to force Condor, to execute simulation and analyse in a row on one worker?
As additional info : I am not using a shared file system
Thanks a lot for your help!
Greetings from Germany