Torque - Revision history

Daniel-schurhoff-de23@rwth-aachen.de at 08:30, 5 September 2019

2019-09-05T08:30:43Z

2019-01-30T11:21:07Z

2018-11-13T12:26:33Z

2018-04-23T05:56:30Z

References

2018-04-23T05:55:39Z

Jobscript Examples

2018-04-23T05:55:19Z

Jobscript Examples

2018-04-20T14:14:21Z

Jobscript Examples

2018-04-20T14:13:04Z

Jobscript Examples

2018-04-20T14:11:22Z

#PBS Usage

2018-04-20T14:09:23Z

#PBS Usage

← Older revision		Revision as of 11:21, 30 January 2019
Line 74:		Line 74:

	For hybrid programs, make sure to disable processor affinity by adding <code>export MV2_ENABLE_AFFINITY=0</code> to your script. Otherwise, all threads will be using the same core.		For hybrid programs, make sure to disable processor affinity by adding <code>export MV2_ENABLE_AFFINITY=0</code> to your script. Otherwise, all threads will be using the same core.
		+
		+	== Array and Chain Jobs ==
		+
		+	<syntaxhighlight lang="zsh">
		+
		+	qsub -W depend=afterok:<Job-ID> <SCRIPT>
		+
		+	</syntaxhighlight>
		+
		+	This requires the Job <Job-ID> to terminate without errors before the submitted script may be run. Other available conditions for chain jobs are
		+
		+	{\| class="wikitable" style="width: 60%;"
		+	\| Condition \|\| Function
		+	\|-
		+	\| afterok:<jobID> \|\| job can start once job <jobID> has terminated without errors
		+	\|-
		+	\| afternotok:<jobID> \|\| job can start once job <jobID> has terminated with errors
		+	\|-
		+	\| afterany:<jobID> \|\| job can start once job <jobID> has terminated
		+	\|}

	== Jobscript Examples ==		== Jobscript Examples ==

@@ Line 1: / Line 1: @@
 == General ==
-Torque is a job [[scheduler]]. To get an overview of the functionality of a scheduler, go [[Scheduler#General|here]].
+Torque is a job [[scheduler]]. To get an overview of the functionality of a scheduler, go [[Scheduler#General|here]] or to the [[Scheduling_Basics|Scheduling Basics]].
 == Job Submission ==

← Older revision		Revision as of 05:56, 23 April 2018
Line 156:		Line 156:

	[http://www.arc.ox.ac.uk/content/torque-job-scheduler Guide to the Torque scheduler]		[http://www.arc.ox.ac.uk/content/torque-job-scheduler Guide to the Torque scheduler]
		+
		+	[https://www.osc.edu/supercomputing/batch-processing-at-osc/job-scripts More jobscript examples and tips]

@@ Line 102: / Line 102: @@
 This job runs the executable "myapp.exe" that has been parallelized with OpenMP using 12 threads.
-<syntax-highlight lang="bash">
+<syntaxhighlight lang="bash">
 #!/usr/bin/bash
@@ Line 123: / Line 123: @@
 ### Execute your application and set OMP_NUM_THREADS for the application run
 OMP_NUM_THREADS=12 myapp.exe
-</syntax-highlight>
+</syntaxhighlight>
 This MPI job will start 8 processes of "myapp.exe" on 2 nodes.
-<syntax-highlight lang="bash">
+<syntaxhighlight lang="bash">
 #!/usr/bin/bash
@@ Line 147: / Line 147: @@
 ### Execute your application and set the "-np" option for the application run
 mpiexec -np 8 myapp.exe
-</syntax-highlight>
+</syntaxhighlight>
 == References ==

@@ Line 78: / Line 78: @@
 This serial job will run a given executable, in this case "myapp.exe".
-<syntaxhighlight lang="zsh">
+<syntaxhighlight lang="bash">
-#!/usr/bin/env zsh
+#!/usr/bin/bash
 ### Job name

@@ Line 71: / Line 71: @@
 |-
 | -l nodes=<num_nodes>:ppn=<num_cores> || specify number of processes to start (one per core), which is num_nodes*num_procs; be careful about your system's architecture and do not request more nodes or cores than available on the machine
-|}
+For hybrid programs, make sure to disable processor affinity by adding <code>export MV2_ENABLE_AFFINITY=0</code> to your script. Otherwise, all threads will be using the same core.
 == Jobscript Examples ==

@@ Line 28: / Line 28: @@
 |-
 | -N <name> || job name
 |-
 | -j oe || both the output and error log will be written to the same log file called <job_name>.o<job_id>
@@ Line 36: / Line 40: @@
 | Parameter || Function
 |-
-| -l walltime=<runlimit> || runtime limit in the format hours:minutes:seconds; once the time specified is up, the job will be killed by the [[scheduler]]
+| -l walltime=<total_time_limit> || time limit (including waiting time in the queue!) in the format hours:minutes:seconds; once the time specified is up, the job will be killed by the [[scheduler]]
 |-
 | -l mem=<memlimit> || memory limit per process as an integer number, followed by a unit, e. g. 400MB
@@ Line 43: / Line 49: @@
 |}
-E-mail notifications:
+Email notifications:
 {| class="wikitable" style="width: 60%;"
 | Parameter || Function
 |-
 | -m a || receive a mail if your job gets aborted
@@ Line 59: / Line 67: @@
 {| class="wikitable" style="width: 60%;"
 | Parameter || Function
+|-
 |}