BinAC/Moab - Revision history

S Behnle: typo

2025-03-14T14:08:25Z

typo

← Older revision		Revision as of 16:08, 14 March 2025
Line 112:		Line 112:
	* [[BinAC/Queues\|bwForCluster BINAC queue settings]]		* [[BinAC/Queues\|bwForCluster BINAC queue settings]]

	With the change from MOAB to Torque, you may have to ~~adept~~ your jobscripts in order to use certain queues:		With the change from MOAB to Torque, you may have to adapt your jobscripts in order to use certain queues:

	* gpu: add <code>:gpu</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''gpu'''</code>		* gpu: add <code>:gpu</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''gpu'''</code>

V Lutz at 12:33, 5 March 2025

2025-03-05T12:33:42Z

← Older revision		Revision as of 14:33, 5 March 2025
Line 114:		Line 114:
	With the change from MOAB to Torque, you may have to adept your jobscripts in order to use certain queues:		With the change from MOAB to Torque, you may have to adept your jobscripts in order to use certain queues:

			* gpu: add <code>:gpu</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''gpu'''</code>
	* short/long/gpu: No change needed
			* short: add <code>:short</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''short'''</code>
			* long: add <code>:long</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''long'''</code>
	* smp: add <code>:smp</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''smp'''</code>		* smp: add <code>:smp</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''smp'''</code>
	* inter: add <code>:inter</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''inter'''</code>		* inter: add <code>:inter</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''inter'''</code>

V Lutz at 10:42, 5 March 2025

2025-03-05T10:42:11Z

← Older revision		Revision as of 12:42, 5 March 2025
Line 77:		Line 77:
	\| -M ''name@uni.de''		\| -M ''name@uni.de''
	\| #PBS -M ''name@uni.de''		\| #PBS -M ''name@uni.de''
	\| Send email to the specified email address "name@uni.de".		\| Send email to the specified email address "name@uni.de".<br>
			Be careful what you wish for, this may generate a lot of emails!
	\|-		\|-
	\|}		\|}

F Bartusch: /* qsub -q queues */

2025-02-03T12:47:22Z

qsub -q queues

← Older revision		Revision as of 14:47, 3 February 2025
Line 110:		Line 110:
	Queue classes define maximum resources such as walltime, nodes and processes per node and partition of the compute system. Note that queue settings of the bwHPC cluster are not '''identical''', but differ due to their different prerequisites, such as HPC performance, scalability and throughput levels. Details can be found here:		Queue classes define maximum resources such as walltime, nodes and processes per node and partition of the compute system. Note that queue settings of the bwHPC cluster are not '''identical''', but differ due to their different prerequisites, such as HPC performance, scalability and throughput levels. Details can be found here:
	* [[BinAC/Queues\|bwForCluster BINAC queue settings]]		* [[BinAC/Queues\|bwForCluster BINAC queue settings]]

			With the change from MOAB to Torque, you may have to adept your jobscripts in order to use certain queues:

			* short/long/gpu: No change needed
			* smp: add <code>:smp</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''smp'''</code>
			* inter: add <code>:inter</code> to the node/proc resource request, e.g.: <code>-l nodes=x:ppn=n:'''inter'''</code>

	=== qsub Examples ===		=== qsub Examples ===

F Bartusch: /* List your jobs and show job details : qstat */

2025-02-03T11:00:29Z

List your jobs and show job details : qstat

← Older revision		Revision as of 13:00, 3 February 2025
Line 349:		Line 349:

	== List your jobs and show job details : qstat ==		== List your jobs and show job details : qstat ==
	Displays information about active, eligible, blocked, and/or recently completed jobs. ~~Since the resource manager is not actually scheduling jobs, the job ordering it displays is not valid. The showq command displays the actual job ordering under the Moab Workload Manager.~~ When used without flags, this command displays all jobs in active, idle, and non-queued states.		Displays information about active, eligible, blocked, and/or recently completed jobs. When used without flags, this command displays all jobs in active, idle, and non-queued states.
	<br>		<br>

F Bartusch: /* Moab Job Control : mjobctl */

2025-02-03T10:58:23Z

Moab Job Control : mjobctl

← Older revision		Revision as of 12:58, 3 February 2025
Line 407:		Line 407:
	job_state = C		job_state = C
	</pre>		</pre>

	== Moab Job Control : mjobctl ==
	The mjobctl command controls various aspects of jobs. It is used to submit, cancel, execute, and checkpoint jobs. It can also display diagnostic information about your own jobs.
	<br>
	=== Canceling own jobs : mjobctl -c ===
	If you want to cancel a job that has been submitted, please do not use the PBS/Torque qdel (n./a.) or the deprecated [[#Canceling own jobs : canceljob\|canceljob]] commands.
	<br>
	<font color=green>Instead, use '''mjobctl -c <jobid>'''. </font>
	<br>
	{\| width=750px class="wikitable"
	! Flag !! Format !! Default !! Description !! Example
	\|- style="vertical-align:top;"
	\| -cl
	\| JobId
	\| (none)
	\| Cancel a job.
	\| see: [[#Example Use of Mjobctl -c\|example use of mjobctl -c]]
	\|}

	=== Example Use of mjobctl -c ===
	Canceling a job on the bwUniCluster
	<pre>
	[...-calc_repo-0]$ qsub bwhpc-fasta-example.moab
	8374426

	$ checkjob 8374426 \| grep ^State
	State: Idle # job is 'Idle'

	$ mjobctl -c 8374426
	job '8374426' cancelled # job is cancelled

	checkjob 8374426 \| grep ^State
	State: Removed # now, job is removed

	$ # my own checkjob wrapper
	cj 8374426
	Job: 8374426 Status: < Removed > Wartezeit: 1m30s Intervall: 30s
	Job 8374426 wurde gelöscht!
	$
	</pre>
	* [[#Checkjob Examples\|checkjob wrapper]]



	=== Other Mjobctl-Options ===
	See also:
	* [http://docs.adaptivecomputing.com/mwm/6-1-9/Content/commands/mjobctl.html Complete list of mjobctl options and parameters].

	<font color=red>Not all of the listed options are available for 'normal' users. Some are for MOAB-admins only.</font>

F Bartusch: /* Canceling own jobs : canceljob */

2025-02-03T10:58:05Z

Canceling own jobs : canceljob

← Older revision		Revision as of 12:58, 3 February 2025
Line 356:		Line 356:
	* For further options of ''qstat'' read the manpage of ''qstat''.		* For further options of ''qstat'' read the manpage of ''qstat''.

	== Canceling own jobs : ~~canceljob~~ ==		== Canceling own jobs : qdel ==
		⚫	The qdel <JobId> command is used to selectively cancel the specified job(s) (active, idle, or non-queued) from the queue.
	<font color=red>Caution: This command is '''deprecated'''. Use [[#Canceling own jobs : mjobctl -c\|mjobctl -c]] instead!</font>
	<br><br>
⚫	The ~~canceljob~~ <JobId> command is used to selectively cancel the specified job(s) (active, idle, or non-queued) from the queue.
	<br><br>		<br><br>
	<font color=red>Note that only '''own jobs''' can be cancelled.</font>		<font color=red>Note that only '''own jobs''' can be cancelled.</font>
	<br>		<br>
	=== Access ===		=== Access ===
	This command can be run by any ~~Moab~~ Administrator and '''by the owner of the job'''.		This command can be run by any Administrator and '''by the owner of the job'''.
	{\| width=750px class="wikitable"		{\| width=750px class="wikitable"
	! Flag !! Name !! Format !! Default !! Description !! Example		! Flag !! Name !! Format !! Default !! Description !! Example
Line 383:		Line 381:
	\|}		\|}

	=== Example Use of ~~Canceljob~~ ===		=== Example Use of qdel ===
	Example use of ~~canceljob~~ run on ~~the bwUniCluster~~		Example use of qdel run on BinAC
	<pre>		<pre>
	[...calc_repo-0]$ qsub bwhpc-fasta-example.~~moab~~		[...calc_repo-0]$ qsub bwhpc-fasta-example.pbs
	8374356 # this is the JobId		8374356 # this is the JobId
	$		$
	$ ~~checkjob~~ 8374356		$ qstat -f 8374356
			(base) [tu_iioba01@login03 samtools]$ qstat -f 11870719
	~~job~~ 8374356		Job Id: 8374356
	AName: fasta36_job
			Job_Name = bwhpc-fasta-example.pbs
	State: Idle
			Job_Owner = tu_iioba01@login03
	Creds: user:kn_pop235844 group:kn_kn account:konstanz class:multinode
			resources_used.cput = 00:00:02
	WallTime: 00:00:00 of 00:10:00
			resources_used.energy_used = 0
	BecameEligible: Fri Jan 15 12:10:53
			resources_used.mem = 1580kb
	SubmitTime: Fri Jan 15 12:10:43
			resources_used.vmem = 56404kb
	(Time Queued Total: 00:00:10 Eligible: 00:00:08)
			resources_used.walltime = 00:00:03
			job_state = R
	[...]		[...]

	$ checkjob 8374356 \| grep ^State:
	State: Idle # state is 'Idle'

	$ # now cancel the job		$ # now cancel the job
	$ ~~canceljob~~ 8374356		$ qdel 8374356
			Terminated
	job '8374356' cancelled

	$ ~~checkjob~~ 8374356 \| grep ~~^State~~:		$ qstat -f 8374356 \| grep state:
			job_state = C
	State: Removed # state turned into 'Removed'
	</pre>		</pre>
	* See: [[#E-Mail notification of mjobctl -c\|E-Mail notification after a job was cancelled/removed]]

	== Moab Job Control : mjobctl ==		== Moab Job Control : mjobctl ==

F Bartusch: /* Blocked job information : checkjob -v */

2025-02-03T10:53:58Z

Blocked job information : checkjob -v

@@ Line 355: / Line 355: @@
 * Show details about a specific job:  <code>qstat -f JOBID </code>
 * For further options of ''qstat'' read the manpage of ''qstat''.
 == Canceling own jobs : canceljob ==

F Bartusch: /* Detailed job information : checkjob */

2025-02-03T10:53:42Z

Detailed job information : checkjob

Show changes

F Bartusch: /* List of your submitted jobs : showq */

2025-02-03T10:53:29Z

List of your submitted jobs : showq

@@ Line 348: / Line 348: @@
 * You do not need an '''exit $exit_code''' in the scripts.
-== List of your submitted jobs : showq ==
+== List your jobs and show job details : qstat ==
 Displays information about active, eligible, blocked, and/or recently completed jobs.  Since the resource manager is not actually scheduling jobs, the job ordering it displays is not valid. The showq command displays the actual job ordering under the Moab Workload Manager.  When used without flags, this command displays all jobs in active, idle, and non-queued states.
 <br>
 == Detailed job information : checkjob ==