[EMAIL PROTECTED] schrieb am 05/05/2008 05:58:00 PM: > > On May 5, 2008, at 10:08 AM, Steve White wrote: > > > What is the right way to do this in a Globus job submission? > > Sounds like a perfect use case for one of the many workflow engines > that are out there (Pegasus, Kepler, etc.). > > You can also do elementary versions of this with metaschedulers such > as GRMS or Gridway.
I think it's not really a good case for metascheduling because it doesn't make sense to separate the (typically very short) setup stage from the actual (long) computation - there are simply no advantages in doing that. If you use a metascheduler, there is risk that your setup job becomes separated in the queue from the computation job by other users' jobs, unnecessarily delaying the whole execution. Another aspect which makes it a rather bad case for metascheduling is the user perception - an MPI job is perceived as *one* entity which should run from start to finish - not a sequence of jobs that may fail independently. Of course, we could and probably should hide the how-many-jobs-are-really-involved technicalities from end users, but life would be simpler if we just didn't have to. Regards, Jan Ploski -- Dipl.-Inform. (FH) Jan Ploski OFFIS Betriebliches Informationsmanagement Escherweg 2 - 26121 Oldenburg - Germany Fon: +49 441 9722 - 184 Fax: +49 441 9722 - 202 E-Mail: [EMAIL PROTECTED] - URL: http://www.offis.de
