Skip to main content

Job accumulation status and estimated waiting time

The figure below shows the status of job accumulation for each queue of Grid Engine and the estimated job wait time (the wait time prediction value of the job from just submitted to start running).

In the line graph below, queueing time is the estimated waiting time of jobs.

  • Job waiting time is estimated by dividing the amount of jobs in the queue by the amount of jobs processed in an hour.
  • The amount of jobs in the queue is shown in blue on the bar graph.
  • The amount of jobs processed per hour is calculated from the 12-hour average of the vertical axis of the line graph (the amount of jobs processed in 10 minutes) multiplied by 6.

The quantity of jobs is calculated as the number of slots occupied by jobs.

job accumlation status
  • Left (the line graph): The amount of jobs per hour that have started running out of the jobs waiting in the queue.
  • Right (the bar chart)
    • Red: amount of jobs currently running
    • Blue: amount of waiting in the queue
    • Green: amount of free resources
    • Pink: amount of jobs in error state

The data is updated every hour.

Reference: Slot definition

Compute node typeDefinition of 1 slot
Thin Type 1a1 CPU core, 8GB memory
Thin Type 1b1 CPU core, 4GB memory
Thin Type 2a1 CPU core, 12GB memory
Thin Type 2b1 CPU core, 16GB memory
Medium1 CPU core, 38.4GB memory
Fat1 CPU core, 47.2GB memory

The number of slots consumed by a job is calculated based on the number of CPU cores requested or the amount of memory, whichever is greater.


  • A job requesting 3 CPU cores and 2 GB memory will request 3 slots.
  • If one CPU core, 12 GB memory is requested in the AMD queue (Thin Type 1a) of the general analysis section, the number of slots requested is calculated to be 2.
当ウェブサイトでは、お客様により良いサービスを提供するため、クッキーを利用しています。 クッキーの利用に同意いただける場合は「同意する」をクリックしてください。