Easy Access to Queue Status

The show_queues command is available on all user-accessible systems at the HPCMP centers. It displays current queuing system information in the following format:

QUEUE INFORMATION:
                |--------------- Limits ---------------| |----------- Counts ----------|
                   Maximum  Queue   User Job-Min Job-Max    Jobs    Jobs   Cores   Cores     Queue
Queue Name        Walltime    Max    Max   Cores   Cores Running Pending Running Pending Type Ena Run
=============== ========== ====== ====== ======= ======= ======= ======= ======= ======= ============
queue #1        hhhh:mm:ss    ###     ###    ###     ###     ###     ###     ###     ###  ***  *   * 
queue #2        hhhh:mm:ss    ###     ###    ###     ###     ###     ###     ###     ###  ***  *   *

NODE INFORMATION:
Node Type             Nodes Available * Cores per Node = Cores Available - Cores Running = Cores Free
===================   ===============   ==============   ===============   =============   ==========
node type #1                #####              ##             ######           ######        ######
node type #2                #####              ##             ######           ######        ######

ADDITIONAL NOTES:
-----------------------------------------------------------------------------------------------------

The fields associated with each column in the QUEUE INFORMATION table are defined as:

  • Queue Name – The name of the queue
  • Maximum Walltime – The maximum time that can be requested by a single job
  • Queue Max – The maximum number of jobs that the queue can run simultaneously
  • User Max – The maximum number of jobs that a single user can run simultaneously
  • Job-Min Cores – The minimum number of cores that a job can request
  • Job-Max Cores – The maximum number of cores that a job can request
  • Jobs Running – The number of jobs currently running
  • Jobs Pending – The number of jobs waiting to run
  • Cores Running – The number of cores in use by all running jobs
  • Cores Pending – The number of cores requested by all jobs waiting to run
  • Queue Type – The queue type (Execution, Routing, ARS, DSP, etc.)
  • Queue Ena – A field signifying (Y/N) whether or not the queue is accepting new job submissions
  • Queue Run – A field signifying (Y/N) whether or not the queue is allowed to run new jobs

The fields associated with each column in the NODE INFORMATION table are defined as:

  • Node Type – A description of the node type (Standard, GPU, Bigmem, Phi, or KNL)
  • Nodes Available – The number of nodes available for use by the batch scheduling system
  • Cores per Node – The number of standard compute cores configured on this node type
  • Cores Available – The number of standard compute cores for this node type that are available for use by the batch scheduling system
  • Cores Running – The number of standard compute cores for this node type that are currently in use by running jobs
  • Cores Free – The number of standard compute cores for this node type that are currently unused and that are available to run jobs

Queue Status Report is of limited use in:

  • Telling a user when their job will run.
  • Choosing a system on which to run jobs.
  • Determining the best queue for a particular job.

Also to note:

  • Some centers may have additional per-queue, per-user, per-group, per-subproject, or other limits that may not be included in this command's output. Users should refer to each center's User Guide and batch scheduling system documentation to find these details.

Back to BC Policy FY10-06.