Reuti wrote: > Am 28.02.2011 um 17:58 schrieb Prentice Bisbal: > >> Hello Everyone. I'm using SGE 6.2u3.One of my users suddenly >> reported getting this error when running Mathematica jobs (which use DRMAA): >> >> >> Java::excptn: A Java exception occurred: >> org.ggf.drmaa.DeniedByDrmException: error: no suitable queues . >> at com.sun.grid.drmaa.SessionImpl.nativeRunJob(Native Method) >> at com.sun.grid.drmaa.SessionImpl.runJob(SessionImpl.java:349) >> >> >> Any ideas what could be causing this? Not sure exactly how to debug >> DRMAA problems, and I can't think if anything I've changed in the past >> few months that would affect this at all. I have a lot of users who used >> gridMathematica heavily in the past, so this definitely worked before... > > This doesn't look like a DRMAA error per se - the job requests something > which can't be satisfied. > > -- Reuti >
I finally figured out the problem, but now I need some advice on coming up with an adequate solution. Here's the cause: A few months ago, I made h_rt a forced resource, and added '-w e' to my global sge_request file so that jobs without h_rt would fail immediately, and the user submitting the job would know something was wrong right away. The problem is that the h_rt isn't 'inherited' by the additional jobs spawned by DRMAA. Using my own .sge_request file to override the global defaults finally allowed me to see this with qstat -j <jobid>: does not request 'forced' resource "h_rt" of queue ... Now how can I fix this? Having the users constantly modify their .sge_request file doesn't seem very practical, and can lead to problems since the .sge_request file would have to have the correct h_rt value when the additional jobs are 'spawned' by DRMAA, not when during the initial job is submitted. gridMathematica is used heavily around here, so this isn't an occasional problem, or something that only 1 or 2 users would have to deal with. I need something permanent and scalable that would not involve any user intervention. I've been digging around the Mathematica documentation, and I don't really see anyway to pass scheduler arguments to the Mathematica jobs. And I'm afraid if I call Wolfram, they'll just say "remove the need for h_rt on your cluster" - Prentice _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
