1just-man-pages/condor_q(1) General Commands Manual just-man-pages/condor_q(1)
2
6 condor_q Display information about jobs in queue
7
9 condor_q [ -help ]
10 condor_q [ -debug ] [ -global ] [ -submitter submitter ] [ -name name ]
12 [ -pool centralmanagerhostname[:portnumber] ] [ -analyze ] [ -better-
13 analyze ] [ -run ] [ -hold ] [ -globus ] [ -goodput ] [ -io ] [ -dag ]
14 [ -long ] [ -xml ] [ -attributes Attr1 [,Attr2 ... ] ] [ -format fmt
15 attr ] [ -cputime ] [ -currentrun ] [ -avgqueuetime ] [ -jobads file ]
16 [ -machineads file ] [ -direct rdbms | quilld | schedd ] [ {cluster |
17 cluster.process | owner | -constraint expression ... } ]
18
20 condor_q displays information about jobs in the Condor job queue. By
21 default, condor_q queries the local job queue but this behavior may be
22 modified by specifying:
23 * the -global option, which queries all job queues in the pool
25 * a schedd name with the -name option, which causes the queue of the
27 named schedd to be queried
28 * a submitter with the -submitter option, which causes all queues of
30 the named submitter to be queried
31 To restrict the display to jobs of interest, a list of zero or more
33 restrictions may be supplied. Each restriction may be one of:
34 * a cluster and a process matches jobs which belong to the specified
36 cluster and have the specified process number
37 * a cluster without a process matches all jobs belonging to the
39 specified cluster
40 * a owner matches all jobs owned by the specified owner
42 * a -constraint expression which matches all jobs that satisfy the
44 specified ClassAd expression. (See section for a discussion of Clas‐
45 sAd expressions.) If no owner restrictions are present in the list,
46 the job matches the restriction list if it matches at least one
47 restriction in the list. If owner restrictions are present, the job
48 matches the list if it matches one of the owner restrictions and at
49 least one non-owner restriction.
50 If the -long option is specified, condor_q displays a long description
52 of the queried jobs by printing the entire job ClassAd. The attributes
53 of the job ClassAd may be displayed by means of the -format option,
54 which displays attributes with a printf(3)format. Multiple -format
55 options may be specified in the option list to display several
56 attributes of the job. If neither -long or -format are specified, con‐
57 dor_q displays a a one line summary of information as follows:
58 ID
60 The cluster/process id of the condor job.
62 OWNER
66 The owner of the job.
68 SUBMITTED
72 The month, day, hour, and minute the job was submitted to the queue.
74 RUN_TIME
78 Wall-clock time accumulated by the job to date in days, hours, min‐
80 utes, and seconds.
81 ST
85 Current status of the job, which varies somewhat according to the
87 job universe and the timing of updates. U = unexpanded (never been
88 run), H = on hold, R = running, I = idle (waiting for a machine to
89 execute on), C = completed, and X = removed.
90 PRI
94 User specified priority of the job, ranges from -20 to +20, with
96 higher numbers corresponding to greater priority.
97 SIZE
101 The virtual image size of the executable in megabytes.
103 CMD
107 The name of the executable.
109 If the -dag option is specified, the OWNER column is replaced with
113 NODENAME for jobs started by Condor DAGMan.
114 If the -run option is specified, the ST, PRI, SIZE, and CMD columns are
116 replaced with:
117 HOST(S)
119 The host where the job is running.
121 If the -globus option is specified, the ST, PRI, SIZE, and CMD columns
125 are replaced with:
126 STATUS
128 The state that Condor believes the job is in. Possible values are
130 PENDING
132 The job is waiting for resources to become available in order to
134 run.
135 ACTIVE
139 The job has received resources, and the application is executing.
141 FAILED
145 The job terminated before completion because of an error, user-
147 triggered cancel, or system-triggered cancel.
148 DONE
152 The job completed successfully.
154 SUSPENDED
158 The job has been suspended. Resources which were allocated for
160 this job may have been released due to a scheduler-specific rea‐
161 son.
162 UNSUBMITTED
166 The job has not been submitted to the scheduler yet, pending the
168 reception of the GLOBUS_GRAM_PROTOCOL_JOB_SIGNAL_COMMIT_REQUEST
169 signal from a client.
170 STAGE_IN
174 The job manager is staging in files, in order to run the job.
176 STAGE_OUT
180 The job manager is staging out files generated by the job.
182 UNKNOWN
186 MANAGER
194 A guess at what remote batch system is running the job. It is a
196 guess, because Condor looks at the Globus jobmanager contact string
197 to attempt identification. If the value is fork, the job is running
198 on the remote host without a jobmanager. Values may also be condor,
199 lsf, or pbs.
200 HOST
204 The host to which the job was submitted.
206 EXECUTABLE
210 The job as specified as the executable in the submit description
212 file.
213 If the -goodput option is specified, the ST, PRI, SIZE, and CMD columns
217 are replaced with:
218 GOODPUT
220 The percentage of RUN_TIME for this job which has been saved in a
222 checkpoint. A low GOODPUT value indicates that the job is failing to
223 checkpoint. If a job has not yet attempted a checkpoint, this column
224 contains [?????].
225 CPU_UTIL
229 The ratio of CPU_TIME to RUN_TIME for checkpointed work. A low
231 CPU_UTIL indicates that the job is not running efficiently, perhaps
232 because it is I/O bound or because the job requires more memory than
233 available on the remote workstations. If the job has not (yet)
234 checkpointed, this column contains [??????].
235 Mb/s
239 The network usage of this job, in Megabits per second of run-time.
241 If the -io option is specified, the ST, PRI, SIZE, and CMD columns are
245 replaced with:
246 READ The total number of bytes the application has read from files
248 and sockets.
249 WRITE The total number of bytes the application has written to files
253 and sockets.
254 SEEK The total number of seek operations the application has per‐
258 formed on files.
259 XPUT The effective throughput (average bytes read and written per
263 second) from the application's point of view.
264 BUFSIZE The maximum number of bytes to be buffered per file.
268 BLOCKSIZE The desired block size for large data transfers.
272 These fields are updated when a job produces a checkpoint or completes.
276 If a job has not yet produced a checkpoint, this information is not
277 available.
278 If the -cputime option is specified, the RUN_TIME column is replaced
280 with:
281 CPU_TIME
283 The remote CPU time accumulated by the job to date (which has been
285 stored in a checkpoint) in days, hours, minutes, and seconds. (If
286 the job is currently running, time accumulated during the current
287 run is not shown. If the job has not produced a checkpoint, this
288 column contains 0+00:00:00.)
289 The -analyze option may be used to determine why certain jobs are not
293 running by performing an analysis on a per machine basis for each
294 machine in the pool. The reasons may vary among failed constraints,
295 insufficient priority, resource owner preferences and prevention of
296 preemption by the PREEMPTION_REQUIREMENTSexpression. If the -long
297 option is specified along with the -analyze option, the reason for
298 failure is displayed on a per machine basis.
299 The -better-analyze option does a more thorough job of determining why
301 jobs are not running than -analyze . There are scalability issues
302 present when run on a pool with a large number of machines, as well as
303 when run to analyze a large number of queued jobs. The -better-analyze
304 option make take an excessively long time to complete in these cases.
305 Therefore, it is recommended to constrain -better-analyze to only ana‐
306 lyze one job at a time.
307
309 -help
310 Get a brief description of the supported options
312 -global
316 Get queues of all the submitters in the system
318 -debug
322 Causes debugging information to be sent to stderr, based on the
324 value of the configuration variable TOOL_DEBUG
325 -submitter submitter
329 List jobs of specific submitter from all the queues in the pool
331 -pool centralmanagerhostname[:portnumber]
335 Use the centralmanagerhostname as the central manager to locate
337 schedds. (The default is the COLLECTOR_HOSTspecified in the configu‐
338 ration file.
339 -analyze
343 Perform an approximate analysis to determine how many resources are
345 available to run the requested jobs. These results are only meaning‐
346 ful for jobs using Condor's matchmaker. This option is never mean‐
347 ingful for Scheduler universe jobs and only meaningful for grid uni‐
348 verse jobs doing matchmaking.
349 -better-analyze
353 Perform a more time-consuming, but potentially more extensive analy‐
355 sis to determine how many resources are available to run the
356 requested jobs.
357 -run
361 Get information about running jobs.
363 -hold
367 Get information about jobs in the hold state. Also displays the time
369 the job was placed into the hold state and the reason why the job
370 was placed in the hold state.
371 -globus
375 Get information only about jobs submitted to grid resources
377 described as gt2 or gt4 .
378 -goodput
382 Display job goodput statistics.
384 -io
388 Display job input/output summaries.
390 -dag
394 Display DAG jobs under their DAGMan.
396 -name name
400 Show only the job queue of the named schedd
402 -long
406 Display job ads in long format
408 -xml
412 Display job ads in xml format. The xml format is fully defined at
414 http://www.cs.wisc.edu/condor/classad/refman/.
415 -attributes Attr1 [,Attr2 ... ]
419 Explicitly list the attributes (by name, and in a comma separated
421 list) which should be displayed when using the -xml or -long
422 options. Limiting the number of attributes increases the efficiency
423 of the query.
424 -format fmt attr
428 Display attribute or expression attr in format fmt . To display the
430 attribute or expression the format must contain a single
431 printf(3)style conversion specifier. Attributes must be from the job
432 ClassAd. Expressions are ClassAd expressions and may refer to
433 attributes in the job ClassAd. If the attribute is not present in a
434 given ClassAd and cannot be parsed as an expression, then the format
435 option will be silently skipped. The conversion specifier must match
436 the type of the attribute or expression. %s is suitable for strings
437 such as Owner, %d for integers such as ClusterId, and %f for float‐
438 ing point numbers such as RemoteWallClockTime. An incorrect format
439 will result in undefined behavior. Do not use more than one conver‐
440 sion specifier in a given format. More than one conversion specifier
441 will result in undefined behavior. To output multiple attributes
442 repeat the -format option once for each desired attribute. Like
443 printf(3)style formats, you can include other text that will be
444 reproduced directly. You can specify a format without any conversion
445 specifiers but you must still give attribute. You can include n to
446 specify a line break.
447 -cputime
451 Instead of wall-clock allocation time (RUN_TIME), display remote CPU
453 time accumulated by the job to date in days, hours, minutes, and
454 seconds. (If the job is currently running, time accumulated during
455 the current run is not shown.)
456 -currentrun
460 Normally, RUN_TIME contains all the time accumulated during the cur‐
462 rent run plus all previous runs. If this option is specified,
463 RUN_TIME only displays the time accumulated so far on this current
464 run.
465 -avgqueuetime
469 Display the average of time spent in the queue, considering all jobs
471 not completed (those that do not have JobStatus == 4or JobStatus ==
472 3.
473 -jobads file
477 Display jobs from a list of ClassAds from a file, instead of the
479 real ClassAds from the condor_schedd daemon. This is most useful for
480 debugging purposes. The ClassAds appear as if condor_q -l is used
481 with the header stripped out.
482 -machineads file
486 When doing analysis, use the machine ads from the file instead of
488 the ones from the condor_collector daemon. This is most useful for
489 debugging purposes. The ClassAds appear as if condor_status -l is
490 used.
491 -direct rdbms | quilld | schedd
495 When the use of Quill is enabled, this option allows a direct query
497 to either the rdbms, the condor_quill daemon, or the condor_schedd
498 daemon for the requested queue information. It also prevents the
499 queue location discovery algorithm from failing over to alternate
500 sources of information for the queue in case of error. It is useful
501 for debugging an installation of Quill. One of the strings rdbms ,
502 quilld , or schedd is required with this option.
503 Restriction list
507 The restriction list may have zero or more items, each of which may
509 be:
510 cluster
512 match all jobs belonging to cluster
514 cluster.proc
518 match all jobs belonging to cluster with a process number of proc
520 -constraint expression
524 match all jobs which match the ClassAd expression constraint
526 A job matches the restriction list if it matches any restriction
528 in the list Additionally, if owner restrictions are supplied, the
529 job matches the list only if it also matches an owner restric‐
530 tion.
531
535 The default output from condor_q is formatted to be human readable, not
536 script readable. In an effort to make the output fit within 80 charac‐
537 ters, values in some fields might be truncated. Furthermore, the Condor
538 Project can (and does) change the formatting of this default output as
539 we see fit. Therefore, any script that is attempting to parse data from
540 condor_q is strongly encouraged to use the -format option (described
541 above, examples given below).
542 Although -analyze provides a very good first approximation, the ana‐
544 lyzer cannot diagnose all possible situations because the analysis is
545 based on instantaneous and local information. Therefore, there are some
546 situations (such as when several submitters are contending for
547 resources, or if the pool is rapidly changing state) which cannot be
548 accurately diagnosed.
549 -goodput , -cputime , and -io are most useful for STANDARD universe
551 jobs, since they rely on values computed when a job checkpoints.
552
554 The -format option provides a way to specify both the job attributes
555 and formatting of those attributes. There must be only one conversion
556 specification per -format option. As an example, to list only Jane
557 Doe's jobs in the queue, choosing to print and format only the owner of
558 the job, the command line arguments for the job, and the process ID of
559 the job:
560 %condor_q -submitter jdoe -format "%s" Owner -format " %s " Args -for‐
562 mat "ProcId = %d0 ProcId
563 jdoe 16386 2800 ProcId = 0
564 jdoe 16386 3000 ProcId = 1
565 jdoe 16386 3200 ProcId = 2
566 jdoe 16386 3400 ProcId = 3
567 jdoe 16386 3600 ProcId = 4
568 jdoe 16386 4200 ProcId = 7
569 To display only the JobID's of Jane Doe's jobs you can use the follow‐
571 ing.
572 %condor_q -submitter jdoe -format "%d." ClusterId -format "%d0 ProcId
574 27.0
575 27.1
576 27.2
577 27.3
578 27.4
579 27.7
580 An example that shows the difference (first set of output) between not
582 using an option to condor_q and (second set of output) using the
583 -globus option:
584 ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD
586 100.0 smith 12/11 13:20 0+00:00:02 R 0 0.0 sleep 10
587 1 jobs; 0 idle, 1 running, 0 held
589 ID OWNER STATUS MANAGER HOST EXECUTABLE
593 100.0 smith ACTIVE fork grid.example.com
594 /bin/sleep
595
597 condor_q will exit with a status value of 0 (zero) upon success, and it
598 will exit with the value 1 (one) upon failure.
599
601 Condor Team, University of Wisconsin-Madison
602
604 Copyright (C) 1990-2009 Condor Team, Computer Sciences Department, Uni‐
605 versity of Wisconsin-Madison, Madison, WI. All Rights Reserved.
606 Licensed under the Apache License, Version 2.0.
607 See the Condor Version 7.4.2 Manual or http://www.condorpro‐
609 ject.org/licensefor additional notices. condor-admin@cs.wisc.edu
610 date just-man-pages/condor_q(1)