Overview
This page details how to run an interactive job from an SSH connection to an ARCC cluster using the salloc command.
...
This page provides step-by-step directions for how to submit a batch job to Slurm from an SSH connection to an ARCC cluster using the sbatch command.
Assuming
Directions start from the assumption you have already ssh’ed into the HPC cluster, or connected to OnDemand and opened the shell access app (). If you have not yet logged into the cluster, start here
...
| Code Block |
|---|
[jsmith@clustername ~]$ salloc -A awesomelab -t 00:10:00 salloc: time_limit: 10.0 salloc: Granted job allocation 123456 salloc: Nodes clusternode-003 are ready for job |
Rejoin an Orphan Spawned Job After Disconnection=
...
Rejoin connection to cluster:
| Code Block |
|---|
[jsmith@localcomputer ~]$ ssh jsmith@medicinebow.arcc.uwyo.edu
###
__ ___ ___ _ ____
/ |/ /__ ____/ (_)____(_)___ ___ / __ )____ _ __
/ /|_/ / _ \/ __ / / ___/ / __ \/ _ \ / __ / __ \ | /| / /
/ / / / __/ /_/ / / /__/ / / / / __/ / /_/ / /_/ / |/ |/ /
/_/ /_/\___/\__,_/_/\___/_/_/ /_/\___/ /_____/\____/|__/|__/
connected to Medicinebow: Welcome Screen & Message of the Day
### |
...
Note job ID by running sacct -u <username>
| Code Block |
|---|
[jsmith@medicinebow.arcc.uwyo.edu ~]$ sacct -u jsmith
JobID JobName Partition Account AllocCPUS State ExitCode
------------ ---------- ---------- ---------- ---------- ---------- --------
123456 interacti+ inv-arcc arcc 1 RUNNING 0:0
123456.int+ interacti+ arcc 1 RUNNING 0:0
123456.ext+ extern arcc 1 RUNNING 0:0 |
...
Note nodes of job by running scontrol to get your headnodescontrol show job 123456 | grep NodeList
...
Interactive jobs can be specified with a number of directives and flags, as shown in this documentation