Job Monitoring: Login and Compute Nodes

Job monitoring is the process of answering the following questions:

  1. What tasks am I currently running codes, processes, and jobs for?

    • A task can be a Slurm job, running Linux processes within that Slurm job, or activities on login nodes.

    • Codes, processes, and jobs on this page are referred to collectively as softwares.

  2. Where are these softwares running?

  3. What resources am I using for a particular task?

  4. Are those resources appropriate for the task?

  5. Am I using those resources efficiently (or as efficiently as possible), or am I wasting resources?

When you understand the monitoring processes presented below, you can answer these questions very quickly.

Slurm jobs and Linux processes are treated separately in each of the links below.

Video Group Number

Video Group Name and Link

Description

1

videos: monitoring login nodes

Videos on how to find and inspect running processes.

2

videos: monitoring compute nodes

Videos on how to find and inspect Slurm jobs and Linux processes.