Ok, so after two jobs start, no others run and they just remain in the queue. Here is what checkjob says on the job that's just sitting there:
[EMAIL PROTECTED] mallet]$ checkjob 274 checking job 274 State: Idle (User: oscartst Group: oscartst) WallTime: 0:00:00 of INFINITY SubmitTime: Thu Jul 1 14:08:29 (Time Queued Total: 0:01:33 Eligible: 0:01:33) Total Tasks: 9 Req[0] TaskCount: 9 Partition: ALL Network: [NONE] Memory >= 0 Disk >= 0 Swap >= 0 Opsys: [NONE] Arch: [NONE] Class: [workq 1] Features: [NONE] IWD: [NONE] Executable: [NONE] QOS: DEFAULT Bypass: 0 StartCount: 0 PartitionMask: [ALL] Reservation '274' ( INFINITY -> INFINITY Duration: INFINITY) PE: 9.00 StartPriority: 1 job cannot run in partition DEFAULT (insufficient idle procs available: 0 < 9) Any ideas from this output? Thanks -jeremy On Wed, 30 Jun 2004, Jeremy Enos wrote: > I can probably help here a bit... > > In PBS, qstat will show all jobs and their state. Keep in mind, that in > typical OSCAR clusters, it is Maui (the job scheduler) which reads PBS's > information about nodes and queues, and instructs PBS on when to run a > given job. If a job isn't running, Maui may be the place to look. > #1 Make sure Maui is running. > #2 Make sure pbs_sched (PBS's included dumbed-down FIFO scheduler) isn't > running and locking the pbs_server port > #3 Use Maui utilities (checkjob,showq?) to investigate and find out why > Maui is or isn't running a given job > > Jeremy > > > At 07:08 PM 6/30/2004, Bernard Li wrote: > >Hey Jeremy: > > > >Is there a way to figure out why PBS isn't running the jobs? > > > >In SGE, there is qstat -j <jid> and it tells you why (queue busy, yadda > >yadda) > > > >Cheers, > > > >Bernard > > > > > -----Original Message----- > > > From: Jeremy Hansen [mailto:[EMAIL PROTECTED] > > > Sent: Wednesday, June 30, 2004 16:57 > > > To: Bernard Li; [EMAIL PROTECTED] > > > Subject: Re: [Oscar-users] Qsub pbs issues > > > > > > > > > Output from pbsnodes -a > > > > > > rlx-back-2-6-10.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-2.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-3.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-4.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-5.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-6.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-7.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-8.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > rlx-back-2-6-9.blahblahblah.net > > > state = job-exclusive > > > np = 2 > > > properties = all > > > ntype = cluster > > > jobs = 0/263.rlx-2-6-1.blahblahblah.net, > > > 1/262.rlx-2-6-1.blahblahblah.net > > > > > > > > > It seems that only a max of two jobs will run simultaneously. > > > > > > Thanks > > > -jeremy > > > > > > > > > On 6/30/04 4:50 PM, "Bernard Li" <[EMAIL PROTECTED]> wrote: > > > > > > > Hey Jeremy: > > > > > > > > Don't remember PBS much but what happens if you do 'pbsnodes -a' ? > > > > > > > > I'll let the guys know that the repository is down - thanks for > > > > letting us know. > > > > > > > > Cheers, > > > > > > > > Bernard > > > > > > > >> -----Original Message----- > > > >> From: Jeremy Hansen [mailto:[EMAIL PROTECTED] > > > >> Sent: Wednesday, June 30, 2004 16:39 > > > >> To: Bernard Li; [EMAIL PROTECTED] > > > >> Subject: Re: [Oscar-users] Qsub pbs issues > > > >> > > > >> Hmm, I'm more then willing to try Torque. It appears the package > > > >> repository is unavailable at the moment. I just don't understand > > > >> though. I'm sure I'm doing something wrong. > > > >> Shouldn't openpbs allocate all available resources? > > > >> Why would it not do this by default? > > > >> > > > >> Thanks > > > >> -jeremy > > > >> > > > >> > > > >> On 6/30/04 4:16 PM, "Bernard Li" <[EMAIL PROTECTED]> wrote: > > > >> > > > >>> Hi Jeremy: > > > >>> > > > >>> Not too sure about OpenPBS mailing-list but there is > > > definitely one > > > >>> for > > > >>> Torque: > > > >>> > > > >>> http://www.supercluster.org/mailing.shtml > > > >>> > > > >>> Torque is basically a 'better' version of OpenPBS with a lot of > > > >>> patches and bug fixes (or is it a complete re-write > > > >> now...?) - so if > > > >>> you want less headaches, I would recommend switching to > > > >> Torque instead. > > > >>> > > > >>> There is a package available for Torque from OPD, and I am sure > > > >>> someone on the list can help you with the switch over... > > > >>> > > > >>> I personally use Sun Grid Engine and loved it. Switched > > > over from > > > >>> OpenPBS long time ago and never looked back ;-) > > > >>> > > > >>> Cheers, > > > >>> > > > >>> Bernard > > > >>> > > > >>>> -----Original Message----- > > > >>>> From: [EMAIL PROTECTED] > > > >>>> [mailto:[EMAIL PROTECTED] On Behalf > > > >> Of Jeremy > > > >>>> Hansen > > > >>>> Sent: Wednesday, June 30, 2004 16:03 > > > >>>> To: [EMAIL PROTECTED] > > > >>>> Subject: [Oscar-users] Qsub pbs issues > > > >>>> > > > >>>> Perhaps this isn't appropriate for this list but I don't know if > > > >>>> OpenPBS even has a list for users. I tried finding one > > > >> but too many > > > >>>> registrations and hassle. > > > >>>> > > > >>>> The issue I'm having, I submit jobs to the queue and they sit in > > > >>>> queue state for no reason even though the nodes are free. > > > >>>> Why doesn't openpbs run the jobs right away? How do I > > > >> force things > > > >>>> to run and allocate nodes immediately? > > > >>>> > > > >>>> -jeremy > > > >>>> > > > >>>> > > > >>>> > > > >>>> > > > >>>> ------------------------------------------------------- > > > >>>> This SF.Net email sponsored by Black Hat Briefings & Training. > > > >>>> Attend Black Hat Briefings & Training, Las Vegas July > > > >> 24-29 - digital > > > >>>> self defense, top technical experts, no vendor pitches, > > > unmatched > > > >>>> networking opportunities. Visit www.blackhat.com > > > >>>> _______________________________________________ > > > >>>> Oscar-users mailing list > > > >>>> [EMAIL PROTECTED] > > > >>>> https://lists.sourceforge.net/lists/listinfo/oscar-users > > > >>>> > > > >>>> > > > >>> > > > >>> > > > >>> ------------------------------------------------------- > > > >>> This SF.Net email sponsored by Black Hat Briefings & Training. > > > >>> Attend Black Hat Briefings & Training, Las Vegas July 24-29 > > > >> - digital > > > >>> self defense, top technical experts, no vendor pitches, unmatched > > > >>> networking opportunities. Visit www.blackhat.com > > > >>> _______________________________________________ > > > >>> Oscar-users mailing list > > > >>> [EMAIL PROTECTED] > > > >>> https://lists.sourceforge.net/lists/listinfo/oscar-users > > > >> > > > >> > > > >> > > > >> > > > > > > > > > > > > > > > > > >------------------------------------------------------- > >This SF.Net email sponsored by Black Hat Briefings & Training. > >Attend Black Hat Briefings & Training, Las Vegas July 24-29 - > >digital self defense, top technical experts, no vendor pitches, > >unmatched networking opportunities. Visit www.blackhat.com > >_______________________________________________ > >Oscar-users mailing list > >[EMAIL PROTECTED] > >https://lists.sourceforge.net/lists/listinfo/oscar-users > > ------------------------------------------------------- This SF.Net email sponsored by Black Hat Briefings & Training. Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com _______________________________________________ Oscar-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/oscar-users
