Drill Down to Jobs

Drill down to obtain information about jobs submitted to an HPC cluster, to a specific queue, or by a specific user.

A colored bar graphically depicts the status of the jobs submitted to an HPC cluster. The below table describes the conditions represented by each color and the associated job states/substates.
Color Condition State/Substate
Gray No Jobs to Report N/A
Red Requires Attention
  • waiting with a substate of 37 (waiting due to file stage in failure)
  • held with a substate of 20 (user or operator hold)
Blue Not Running
  • queued
  • suspended
  • suspended due to workstation becoming busy
  • waiting with any substate except 37
  • held with any substate except 20
Green Running
  • running
  • exiting
  • transitioning

Drill Down to Jobs Submitted to an HPC Cluster

Drill-down to jobs that have been submitted to an HPC cluster.

Drill-down to see all jobs that have been submitted to an HPC cluster or jobs that have been submitted to a cluster that are in specific state.

  1. Click the Monitor tab.
  2. Click the Jobs tab from the PBS Professional menu located on the left-hand side of the web page.
  3. Choose one of the following options:
    • To view all jobs submitted to the HPC cluster, click the number under the Jobs column.
    • To view jobs submitted to the HPC cluster that are in a specific state, click the appropriate colored area of the Status bar.

    Drill-down to HPC jobs
    Figure 1. Drill-down to HPC Jobs
    A list matching the selected filtering requirements is presented with each row representing a job submitted to the HPC cluster.

Drill Down to Jobs Directed to a Queue

Drill-down to jobs that have been directed to a particular queue.

Drill-down to see all jobs that have been directed to particular queue or to jobs directed to a queue that are in specific state.

  1. Click the Monitor tab.
  2. Click the Jobs tab from the PBS Professional menu located on the left-hand side of the web page.
  3. Expand the HPC cluster by clicking on its name.
    A list of queues is displayed.
  4. Choose one of the following options:
    • To view all jobs directed to the queue, click the number under the Jobs column.
    • To view jobs directed to a queue that are in a specific state, click the appropriate colored area of the Status bar.

    Drill-down to Jobs Submitted to a Queue
    Figure 2. Drill-down to Jobs Submitted to a Queue
    A list matching the selected filtering requirements is presented with each row representing a job submitted to that queue.

Drill Down to Jobs Directed to a Queue by User

Drill-down to jobs that have been directed to a particular queue by user.

Drill-down by queue to see all jobs that have been submitted by a user or jobs submitted by a user that are in specific state.

  1. Click the Monitor tab.
  2. Click the Jobs tab from the PBS Professional menu located on the left-hand side of the web page.
  3. Expand the HPC cluster by clicking on its name.
    A list of queues is displayed.
  4. Expand a queue by clicking on its name.
    A list of users who have submitted jobs is displayed.
  5. Choose one of the following options:
    • To view all jobs submitted by that user, click the number under the Jobs column.
    • To view jobs submitted by that user that are in a specific state, click the appropriate colored area of the Status bar.

    Drill-down to jobs submitted by a user
    Figure 3. Drill-down to Jobs Submitted by a User
    A list matching the selected filtering requirements is presented with each row representing a job submitted by that user.

Understanding the Job Detail View

A detailed view of HPC jobs including state, queue, user, number of CPUs requested, and the job progress.

By default, the Job Detail view displays all jobs that have been submitted to the HPC, but have not yet finished. Use this view to drill-down and investigate issues with jobs and perform job management tasks such as placing a job on hold, moving a job to another queue, suspending a job or deleting a job. Each job is represented by a row containing the following information:

  • job identifier
  • the state of the job
  • job name
  • user who submitted the job
  • queue to which the job was submitted
  • number of CPUs requested
  • progress of the job if walltime is specified at job submission
  • comments added by the WLM administrator or the WLM itself about the job

Job Arrays

Job arrays offer a mechanism for submitting and managing collections of similar jobs quickly and easily. Each job in a job array is called a subjob. The job identifier assigned to a job array has the format jobid[].servername. The job identifiers of the subjobs in the same job array are the same except for their indices. Each subjob has a unique index.


Job Details
Figure 4. Job Details