search

NVIDIA / pyxis

Container plugin for Slurm Workload Manager
Apache License 2.0
263 stars 28 forks source link

Fail to restart slurmd server after tweaking my PMIx configuration through systemd #98

Closed inspurasc closed 1 year ago

inspurasc commented 1 year ago

hello,

I set pmix follow the steps as : https://github.com/NVIDIA/pyxis/wiki/Setup#slurmd-configuration My slurmd location in /opt/slurm/20.11.9/sbin/slurmd . And I don't found /etc/default/slurmd file , so i add it m0anually. But I fail to restart the slurmd service after the seting.

flx42 commented 1 year ago

What is the error message?

inspurasc commented 1 year ago

It seems the slurmd has started, but it stopped again for some reason。 This is the slurmd.log output.

image

And I faild to create the /etc/default/slurmd file. It prompted "/etc/sysconfig/slurmd" E212: Can't open file for writing". I have used the root permission

Please give me some suggestions, thanks

inspurasc commented 1 year ago

Please provide some suggestions for above issue, thanks.

flx42 commented 1 year ago

There are no errors visible in your log, try increasing slurmd verbosity perhaps.