Hi Daneil I am in he process of identifying the problem It's peculiar One advice I can give you is to remove the mem parameter from the submit script. The fact is that while scheduling the job maui only checks for resource requirement in the initial stage.After the job is scheduled it can use whatever resource it can in terms of memory and disk space.Your job will surely get executed.Although other scheduled jobs will have less resources available to them If I get time I will surely try it over my cluster and will convey the result to you as soon as possible. Bye One more thing:While replying do include maui user's list so that other's can also ponder over the prolem.That's why the user's list is made :) Bye again.Will get in touch with you very soon with a positive reply
On 5/15/07, Daniel Boone <[EMAIL PROTECTED]> wrote:Hi
I tried some new parameters. print server output of qmgr ---------------- create queue batch set queue batch queue_type = Execution set queue batch resources_default.mem = 2000mb set queue batch resources_default.nodes = 1 set queue batch resources_default.pvmem = 16000mb set queue batch resources_default.walltime = 06:00:00 set queue batch enabled = True set queue batch started = True # # Set server attributes. # set server scheduling = True set server managers = [EMAIL PROTECTED] set server operators = [EMAIL PROTECTED] set server default_queue = batch set server log_events = 511 set server mail_from = adm set server scheduler_iteration = 600 set server node_check_rate = 150 set server tcp_timeout = 6 set server pbs_version = 2.1.8 ---------------------- checkjob output: ---------------------- checking job 90 (RM job '90.em-research00') State: Idle EState: Deferred Creds: user:abaqus group:users class:batch qos:DEFAULT WallTime: 00:00:00 of 5:00:00 SubmitTime: Tue May 15 11:59:03 (Time Queued Total: 1:58:17 Eligible: 00:00:00) Total Tasks: 4 Req[0] TaskCount: 4 Partition: ALL Network: [NONE] Memory >= 0 Disk >= 0 Swap >= 15G Opsys: [NONE] Arch: [NONE] Features: [NONE] Exec: '' ExecSize: 0 ImageSize: 0 Dedicated Resources Per Task: PROCS: 1 MEM: 250M SWAP: 15G NodeAccess: SHARED TasksPerNode: 2 NodeCount: 2 IWD: [NONE] Executable: [NONE] Bypass: 0 StartCount: 0 PartitionMask: [ALL] SystemQueueTime: Tue May 15 13:00:06 Flags: RESTARTABLE job is deferred. Reason: NoResources (cannot create reservation for job '90' (intital reservation attempt) ) Holds: Defer (hold reason: NoResources) PE: 6.07 StartPriority: 57 cannot select job 90 for partition DEFAULT (job hold active) ------------------- pbs-script: ------------------- #!/bin/bash #PBS -l nodes=2:ppn=2 #PBS -l walltime=05:00:00 #PBS -l mem=1000mb #PBS -l vmem=7000mb #PBS -j oe #PBS -M [EMAIL PROTECTED] #PBS -m bae # Go to the directory from which you submitted the job mkdir $PBS_O_WORKDIR string="$PBS_O_WORKDIR/plus2gb.inp" scp 10.1.0.52:$string $PBS_O_WORKDIR cd $PBS_O_WORKDIR #module load abaqus # /Apps/abaqus/Commands/abaqus job=plus2gb queue=abaqus4cpu input=Standard_plus2gbyte.inp cpus=4 --------------------------- abaqus environment file. -------------------------- import os os.environ['LAMRSH'] = 'ssh' max_cpus=6 mp_host_list=[['em-research00',3],['10.1.0.97',2]] run_mode = BATCH scratch = "/home/abaqus" queue_name=["cpu","abaqus4cpu"] queue_cmd="qsub -r n -q batch -S /bin/bash -V -l nodes=1:ppn=1 %S" cpu="qsub -r n -q batch -S /bin/bash -V -l nodes=1:ppn=2 %S" abaqus4cpu="qsub -r n -q batch -S /bin/bash -V -l nodes=2:ppn=2 %S" pre_memory = "3000 mb" standard_memory = "7000 mb" --------------------------- but still no changes thanks for al the help until now. rishi pathak schreef: > Also try in your job script file > PBS -l pvmem=<amount of virtual memory> > > On 5/15/07, *rishi pathak* <[EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>> wrote: > > I did not see any specific queue in th submit script > have you specified the following for the queue you are using > > resources_default.mem #available ram > resources_default.pvmem #virtual memory > > > > > > On 5/15/07, *Daniel Boone* <[EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>> wrote: > > Hi > > I need to use the swap. I know I don't have enough RAM, but > the job must > be able to run. Even if it swaps a lot. > Time is not an issue here. > On 1 machine the job uses about 7.4GB swap. We don't have any > other > machines with more RAM to run it on. > Otherwise the other option is to run the job outside > torque/maui, but I > rather don't do that. > > Can some tell me how to read the checkjob -v output, because I > don't > understand how to find errors in it. > > rishi pathak schreef: > > Hi > > system memory(RAM) available to per process is less than the > requested > > amount > > It is not considering swap as an extention of RAM > > Try with reduced system memory > > > > > > > > On 5/14/07, *Daniel Boone* <[EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]> > > <mailto: [EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>>> wrote: > > > > Hi > > > > I'm having the following problem. When I submit a very > > memory-intensive(most swap) job, the job doesn't want to > start. > > It gives the error: cannot select job 62 for partition > DEFAULT > > (job hold > > active) > > But I don't understand what the error means. > > > > I run torque 2.1.8 with maui maui-3.2.6p19 > > > > checkjob -v returns the following: > > ------------------- > > checking job 62 (RM job '62.em-research00') > > > > State: Idle EState: Deferred > > Creds: user:abaqus group:users class:batch qos:DEFAULT > > WallTime: 00:00:00 of 6:00:00 > > SubmitTime: Mon May 14 14:13:41 > > (Time Queued Total: 1:53:39 Eligible: 00:00:00) > > > > Total Tasks: 4 > > > > Req[0] TaskCount: 4 Partition: ALL > > Network: [NONE] Memory >= 0 Disk >= 0 Swap >= 0 > > Opsys: [NONE] Arch: [NONE] Features: [NONE] > > Exec: '' ExecSize: 0 ImageSize: 0 > > Dedicated Resources Per Task: PROCS: 1 MEM: 3875M > > NodeAccess: SHARED > > TasksPerNode: 2 NodeCount: 2 > > > > > > IWD: [NONE] Executable: [NONE] > > Bypass: 0 StartCount: 0 > > PartitionMask: [ALL] > > SystemQueueTime: Mon May 14 15:14:13 > > > > Flags: RESTARTABLE > > > > job is deferred. Reason: NoResources (cannot create > reservation for > > job '62' (intital reservation attempt) > > ) > > Holds: Defer (hold reason: NoResources) > > PE: 19.27 StartPriority: 53 > > cannot select job 62 for partition DEFAULT (job hold active) > > ------------------------ > > checknode of the two nodes:checking node em-research00 > > ------------ > > State: Idle (in current state for 2:31:21) > > Configured Resources: PROCS: 3 MEM: 2010M SWAP: > 33G DISK: 72G > > > > > > Utilized Resources: DISK: 9907M > > Dedicated Resources: [NONE] > > Opsys: linux Arch: [NONE] > > Speed: 1.00 Load: 0.000 > > Network: [DEFAULT] > > Features: [F] > > Attributes: [Batch] > > Classes: [batch 3:3] > > > > Total Time: 2:29:18 Up: 2:29:18 (100.00%) Active: > 00:00:00 (0.00% ) > > > > Reservations: > > NOTE: no reservations on node > > > > -------------------- > > State: Idle (in current state for 2:31:52) > > Configured Resources: PROCS: 2 MEM: 2012M SWAP: > 17G DISK: 35G > > Utilized Resources: DISK: 24G > > Dedicated Resources: [NONE] > > Opsys: linux Arch: [NONE] > > Speed: 1.00 Load: 0.590 > > Network: [DEFAULT] > > Features: [NONE] > > Attributes: [Batch] > > Classes: [batch 2:2] > > > > Total Time: 2:29:49 Up: 2:29:49 ( 100.00%) Active: > 00:00:00 ( 0.00%) > > > > Reservations: > > NOTE: no reservations on node > > ----------------- > > The pbs scipt I'm using: > > #!/bin/bash > > #PBS -l nodes=2:ppn=2 > > #PBS -l walltime=06:00:00 > > #PBS -l mem=15500mb > > #PBS -j oe > > # Go to the directory from which you submitted the job > > mkdir $PBS_O_WORKDIR > > string="$PBS_O_WORKDIR/plus2gb.inp" > > scp 10.1.0.52:$string $PBS_O_WORKDIR > > #scp 10.1.0.52:$PBS_O_WORKDIR'/'$PBS_JOBNAME ./ > > cd $PBS_O_WORKDIR > > #module load abaqus > > # > > /Apps/abaqus/Commands/abaqus job=plus2gb queue=cpu2 > > input=Standard_plus2gbyte.inp cpus=4 mem=15000mb > > --------------------------- > > If you need some extra info please let me know. > > > > Thank you > > > > _______________________________________________ > > mauiusers mailing list > > [email protected] > <mailto:[email protected]> <mailto: > [email protected] <mailto:[email protected]>> > > http://www.supercluster.org/mailman/listinfo/mauiusers > > > > > > > > > > -- > > Regards-- > > Rishi Pathak > > National PARAM Supercomputing Facility > > Center for Development of Advanced Computing(C-DAC) > > Pune University Campus,Ganesh Khind Road > > Pune-Maharastra > > > > > -- > Regards-- > Rishi Pathak > National PARAM Supercomputing Facility > Center for Development of Advanced Computing(C-DAC) > Pune University Campus,Ganesh Khind Road > Pune-Maharastra > > > > > -- > Regards-- > Rishi Pathak > National PARAM Supercomputing Facility > Center for Development of Advanced Computing(C-DAC) > Pune University Campus,Ganesh Khind Road > Pune-Maharastra
-- Regards-- Rishi Pathak National PARAM Supercomputing Facility Center for Development of Advanced Computing(C-DAC) Pune University Campus,Ganesh Khind Road Pune-Maharastra
_______________________________________________ mauiusers mailing list [email protected] http://www.supercluster.org/mailman/listinfo/mauiusers
