List state of all compute nodes

The cluster is running SLURM job scheduler, responsible for allocating the resources requested for batch or jobs. A list of commands that can be used:
# sacct (displays job accounting information about active or completed jobs)
# salloc (allocates resources for a job in real time)
# sbatch (submits a job script for execution)
# scancel (cancels a pending or running job)
# sinfo -N -l (reports the state of partitions and nodes)
# squeue (reports the state of jobs)
# srun (submits a job for real time execution e.g. for interactive jobs)
# scontrol (For viewing and modifying Slurm configuration and states)

To update the status of one node (more info about scontrol command:
# scontrol update nodename=compute003 state=idle

SLURM submit script

-- Last updated by AlexeyLyapin - 16 Oct 2017

Topic attachments
I Attachment History Action Size Date WhoSorted ascending Comment
Unix shell scriptsh r1 manage 0.5 K 16 Oct 2017 - 09:53 AlexeyLyapin SLURM submit script

This topic: Public > WebHome > HpcFarm > HpcSLURM
Topic revision: r4 - 16 Oct 2017 - AlexeyLyapin
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2023 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding RHUL Physics Department TWiki? Send feedback