Setting up slurm on a cluster

Question

My IT admin has setup a cluster with 3 nodes, which is administered via Windows server. VMs are hosted via Hyper-V, including an Ubuntu VM to which a substantial portion of the cluster's resources have been allocated. Does anyone have any experience/success with setting up slurm for job scheduling on a Linux VM hosted on Hyper-V? Existing examples I've found online for setting up slurm don't seem to be working, with a litany of errors. If anyone can point me to some general steps used to setup slurm, would be appreciated.

Edits: have added some errors I am seeing below...

Errors I am getting include: when I scontrol ping: ** RESTORE SLURMCTLD DAEMON TO SERVICE **

and for sinfo --Node --long slurm_load_partitions: Unable to contact slurm controller (connect failure)

scontrol show daemons does work, returning: slurmctld slurmd

However, scontrol show config gives me: slurm_load_ctl_conf error: Unable to contact slurm controller (connect failure)

I have successfully setup slurm just recently on other machines and have never had any such errors. Have tried googling these, no luck so far re any suggestions I have seen on forums.

The errors you encounter would be helpful. The fact that it's a VM should not matter. — Gerald Schneider, Jan 19 '22 at 09:02
That it's on a VM should make zero difference to whatever issues you're facing — Chopper3, Jan 19 '22 at 10:12
Thanks a lot; looking into the log file helped; everything is up and running now. Basically, there were a number of permissions' issues, has to chmod a couple of files and folder locations and now everything is working fine. — ml_white_belt, Jan 19 '22 at 19:55

Setting up slurm on a cluster

0 Answers0