Submitting jobs to the cluster

To manage all the compute jobs on the cluster, we use Slurm. Slurm is a powerful workload manager commonly used in High-Performance Computing (HPC) environments. It is used to allocate resources, such as CPUs and GPUs, to users for a specified duration, usually until the completion of a job. To ensure equitable access to resources, Slurm manages a job queue.

Submitting batch jobs

To submit and monitor batch jobs, have a look at our page on running jobs

Interactive jobs

Sometimes it can be useful to have interactive jobs, for example when debugging a script. Have a look at our page on interactive jobs to learn how.

Monitoring jobs

You can monitor your jobs to keep an eye on resource usage, see if they have finished, etc. We explain how to do this on our monitoring page

Getting project information

First you need to figure out in which compute project that your jobs should be run. Each compute project has a given computer time allocation that is measured in "core-hours per month" (which is treated as a rolling 30 day window). This information can be obtained from the projinfo program. Run projinfo on the login node and study the output:

[emilia@vera ~]$ projinfo -u emilia
Running as user: emilia
Project                 Used[h]        Allocated[h]       Queue
   User
---------------------------------------------------------------
NAISS001-23-456        61734.61*              55000       vera
   emilia              12308.91*
   f38anlo             49104.36* <-- star means we are over 100% usage
   emil                  321.34*     which means this project has lowered priority
---------------------------------------------------------------
NAISS1234-5-67             7.12              100000       vera
   fantomen                7.12

In this example we are logged in as the user emilia that is a member of two projects, NAISS001-23-456 and NAISS1234-5-67 that have monthly time allocations of 55000 and 100000 core-hours/month respectively. The Used column shows the total use since the beginning of the 30-day window in total, and divided up on different members of each project. As most of the current months allocation in the NAISS001-23-456 project already has been used by the user f38anlo we decide to submit our jobs to the project NAISS1234-5-67 with almost no previous usage.

Note

The projinfo tool only shows information about compute projects (like core-hours used). To check storage usage and paths, use the C3SE_quota command instead. You can find more details on our storage and file system page.