Command |
Description |
|---|---|
|
Show partition and node status. |
|
Show your running or pending jobs. |
|
Submit a batch job script. |
|
Cancel a specific job. |
|
Show detailed job information. |
|
Show job accounting information. |
Resource Type |
Slurm Option |
Example |
|---|---|---|
Partition |
|
|
Time limit |
|
|
CPUs per task (cores on |
|
|
Nodes |
|
|
Tasks (MPI ranks) |
|
|
Whole cores only ( |
|
|
Memory per node |
|
|
Memory per CPU |
|
|
GPUs |
|
|
Job array |
|
|
You are in the wrong directory. cd ~/hpc_examples and use the path shown in the instructions.
The job hasn’t started yet (check squeue -u $USER), or you replaced <jobid> wrongly. ls -t *.out | head shows the newest output files.
The last column of squeue gives the reason. Resources and Priority mean it is waiting its turn.
Check how busy the partition is with sinfo.
Reduce the CPUs, memory or time you request, or use a different partition.
Increase --mem.
Check what a finished job really used with sacct -j <jobid> --format=JobID,MaxRSS.
Did you run the module unuse / module use lines first? See Loading Software with Modules.
Use module spider <software> to search, and include the version number.
Ensure the SIF file has correct permissions.
Try pulling the image again.
Check whether the container needs specific bind mounts.
System Status: Check https://hpc.cloudcix.com for announcements.
Documentation: This guide and the Open OnDemand help pages.
Support: Contact CloudCIX support at support@cloudcix.com.
Slurm Official Documentation: https://slurm.schedmd.com/documentation.html
Apptainer Documentation: https://apptainer.org/docs/
Conda User Guide: https://docs.conda.io/projects/conda/en/latest/user-guide/