Tracking jobs in Slurm

Here is some useful code for tracking current and historical jobs on a High Performance Computing system.

The Slurm Workload Manager, formerly known as Simple Linux Utility for Resource Management, or simply Slurm, is a free and open-source job scheduler for Linux and Unix-like kernels, used by many of the world’s supercomputers and computer clusters.

I use Slurm on a daily basis to check the status of my submitted tasks, and to view historical jobs.

Viewing Current Tasks

To view current tasks, I use the following code:

squeue -u $USER -o "%.18i %.9P %.35j %.8u %.8T %.10M %.10l %.6D %20R %80Z"

This will display the current task, including:

• Job ID
• Partition
• Name
• User
• State
• Time
• Time Limit
• Nodes
• Nodelist and Reason
• Working Directory

Viewing Historical Tasks

To view historical tasks, use the following code:

sacct -u $USER \
  --starttime=$(date -d "3 months ago" +%Y-%m-%d) \
  --format=JobID,JobName%30,Partition,AllocCPUS,State,ExitCode,Elapsed,Start,End \
  -X

This will display the following information for all jobs over the past 3 months:

• Job ID
• Job Name
• Partition
• Number of Allocated CPUs
• State
• Exit Code
• Time Elapsed
• Start Time
• End Time

If a longer timespan is required, adjust the starttime field.

Have Your Say

Δ