Moab Workload Manager

checkjob

(Check Job)

Synopsis

checkjob [-A] [-l policylevel] [-n nodeid] [-q qosid] [-r reservationid] [-v] [--flags=future] jobid 

Overview

   checkjob displays detailed job state information and diagnostic output for a specified job.  Detailed information is available for queued, blocked, active, and recently completed jobs.

Access

   This command can be run by level 1-3 Moab administrators for any job.  Also, end users can use checkjob to view the status of their own jobs.

Arguments

-A (Attribute-Value pair)
 
---
Provides output in the form of parsable Attribute-Value pairs.
> checkjob -A 6235

Moab will display job information in the following format: <ATTRIBUTE>=<VALUE>;.

   
--flags
--flags=future
---
Evaluates future eligibility of job (ignore current resource state and usage limitations).
> checkjob -v --flags=future 6235

Display reasons why idle job is blocked ignoring node state and current node utilization constraints.

   
-l (Policy level)
<POLICYLEVEL>

HARD, SOFT, or OFF
---
Reports job start eligibility subject to specified throttling policy level.
> checkjob -l SOFT 6235
> checkjob -l HARD 6235
   
-n (NodeID)
<NODEID>
---
Checks job access to specified node and preemption status with regards to jobs located on that node.
> checkjob -n node113 6235
   
-q (QoS)
<QOSID>
---
Checks job access to specified QoS <QOSID>.
> checkjob -q special 6235
   
-r (Reservation)
<RSVID>
---
Checks job access to specified reservation <RSVID>.
> checkjob -r orion.1 6235
   
-v (Verbose)
 
N/A
Sets verbose mode.
> checkjob -v 6235

Details

   This command allows any Moab administrator to check the detailed status and resource requirements of a active, queued, or recently completed job.  Additionally, this command performs numerous diagnostic checks and determines if and where the job could potentially run.  Diagnostic checks include policy violations, reservation constraints, preemption status, and job to resource mapping.  If a job cannot run, a text reason is provided along with a summary of how many nodes are and are not available.  If the -v flag is specified, a node by node summary of resource availability will be displayed for idle jobs.

Job Eligibility

   If a job cannot run, a text reason is provided along with a summary of how many nodes are and are not available.  If the -v flag is specified, a node by node summary of resource availability will be displayed for idle jobs.  For job level eligibility issues, one of the following reasons will be given:
 
Reason Description
one or more job holds are currently in place
there are currently not adequate processor resources available to start the job
adequate idle processors are available but these do not meet job requirements
job has specified a minimum start date which is still in the future
job is in an unexpected state
job is not in the idle state
job depends on another job reaching a certain state
job start is prevented by a throttling policy 

   If a job cannot run on a particular node, one of the following 'per node' reasons will be given:
 
Node does not allow required job class/queue
Node does not possess required processors
Node does not possess required local disk
Node does not possess required node features
Node does not possess required real memory
Node does not possess required network interface 
Node is not Idle or Running

Reservation Access

   The -r flag can be used to provide detailed information about job access to a specific reservation

Preemption Status

   If a job is marked as a preemptor and the -v and -n flags are specified, checkjob will perform a job by job analysis for all jobs on the specified node to determine if they can be preempted.

Output

   The checkjob command displays the following job attributes:
 
Attribute Value Description
<STRING> Name of account associated with job
[[[DD:]HH:]MM:]SS Length of time job actually ran.

Note This info is only displayed in simulation mode.

Square bracket delimited list of node and processor ids List of nodes and processors allocated to job
<STRING> Node architecture required by job
square bracket delimited list of job attributes Job Attributes (i.e. [BACKFILL][BENCHMARK][PREEMPTEE])
<FLOAT> Average load balance for a job
<FLOAT>  
<INTEGER> Number of times a lower priority job with a later submit time ran before the job
[<CLASS NAME> <CLASS COUNT>] Name of class/queue required by job and number of class initiators required per task.
<INTEGER>  
<INTEGER> Amount of local disk required by job (in MB)
[[[DD:]HH:]MM:]SS The scheduler's estimated walltime.

Note In simulation mode, it is the actual walltime.

<INTEGER> Size of job executable (in MB)
<STRING> Name of command to run
Square bracket delimited list of <STRING>s Node features required by job
   
<STRING> Name of Unix® group associated with job
Zero or more of User, System, and Batch Types of job holds currently applied to job
<INTEGER> Size of job data (in MB)
<DIR> Directory to run the executable in
<INTEGER> Amount of real memory required per node (in MB)
<FLOAT>  
<STRING  Type of network adapter required by job
   
<INTEGER  Number of nodes required by job
<STRING  Node operating system required by job
ALL or colon delimited list of partitions List of partitions the job has access to
<FLOAT> Number of processor-equivalents requested by job
<STRING> Quality of Service associated with job
<RSVID  ( <TIME1  -  <TIME2> Duration: <TIME3>) RESID specifies the reservation id, TIME1 is the relative start time, TIME2 the relative end time, TIME3 the duration of the reservation
[<INTEGER>] TaskCount: <INTEGER> Partition: <partition> A job requirement for a single type of resource followed by the number of tasks instances required and the appropriate partition
<INTEGER> Number of times job has been started by Moab
<INTEGER> Start priority of job
<TIME> Time job was started by the resource management system
One of Idle, Starting, Running, etc Current Job State
<TIME> Time job was submitted to resource management system
<INTEGER> Amount of swap disk required by job (in MB)
Square bracket delimited list of nodes  
<INTEGER> Number of nodes requested by job
<INTEGER> Number of tasks requested by job
<STRING> Name of user submitting job
<FLOAT>  
[[[DD:]HH:]MM:]SS of [[[DD:]HH:]MM:]SS Length of time job has been running out of the specified limit
In the above table, fields marked with an asterisk (*) are only displayed when set or when the -v flag is specified.
 

Example 1

  Check job 717

> checkjob 717
job 717

State: Idle
Creds:  user:jacksond  group:jacksond  class:batch
WallTime: 00:00:00 of 00:01:40
SubmitTime: Mon Aug 15 20:49:41
  (Time Queued  Total: 3:12:23:13  Eligible: 3:12:23:11)

TerminationDate:   INFINITY  Sat Oct 24 06:26:40
Total Tasks: 1

Req[0]  TaskCount: 1  Partition: ALL
Network: ---  Memory >= 0  Disk >= 0  Swap >= 0
Opsys: ---  Arch: ---  Features: ---


IWD:            /home/jacksond/moab/moab-4.2.3
Executable:     STDIN
Flags:          RESTARTABLE,NORMSTART
StartPriority:  5063
Reservation '717' (  INFINITY ->   INFINITY  Duration: 00:01:40)
Note:  job cannot run in partition base (idle procs do not meet requirements : 0 of 1 procs found)
idle procs:   4  feasible procs:   0

Rejection Reasons: [State        :    3][ReserveTime  :    1]

cannot select job 717 for partition GM (partition GM does not support requested class batch)

Note The example job cannot be started for two different reasons.
  • It is temporarily blocked from partition base because of node state and node reservation conflicts.
  • It is permanently blocked from partition GM because the requested class batch is not supported in that partition.

See Also

  • Moab Client Installation - explains how to distribute this command to client nodes
  • mdiag -j command - display additional detailed information regarding jobs
  • showq command - showq high-level job summaries
  • JOBCPURGETIME parameter - specify how long information regarding completed jobs is maintained
  • diagnosing job preemption