search

nateGeorge / slurm_gpu_ubuntu

Instructions for setting up a SLURM cluster using Ubuntu 18.04.3 with GPUs.
MIT License
131 stars 37 forks source link

Couldn't find the specified plugin name for cgroup/v2 looking at all files #7

Open shubhamk8597 opened 10 months ago

shubhamk8597 commented 10 months ago

Thank you @nateGeorge for the detailed repo. This has helped me a lot for the initial setup.

Unfortunately I am stuck with cgroupv2.

When I try to restart slurm I get the following error in slurmd.log

[2023-10-16T10:42:01.138] error: Couldn't find the specified plugin name for cgroup/v2 looking at all files [2023-10-16T10:42:01.139] error: cannot find cgroup plugin for cgroup/v2 [2023-10-16T10:42:01.139] error: cannot create cgroup context for cgroup/v2 [2023-10-16T10:42:01.139] error: Unable to initialize cgroup plugin [2023-10-16T10:42:01.139] error: slurmd initialization failed

I looked it up and found the following resources. https://slurm.schedmd.com/cgroup_v2.html

I installed the required packages at the bottom of the page but no luck. I am using slurm-23.02.5 and Ubuntu 20.04.6 Do you have any tips as to where to look? Thank you for your time.

fatemesbati commented 7 months ago

hi how did you fix it? in my controller node i have no error, but in my compute nodes i have exact same error

ezratamu commented 5 months ago

Would love a solution for this

nateGeorge commented 4 months ago

Sorry, I haven't touched this in years and won't be any help. Maybe ask this on the new repo for ubuntu 22: https://github.com/lopentusska/slurm_ubuntu_gpu_cluster