Join Log in
Score:0
joshuaclst avatar Ubuntu

Failed to start Slurm node daemon

au flag
joshuaclst

I want to install slurm to manage properly my DIY cluster. I want to use the cluster (HPC) for parallel simulations. I have 3 nodes (1 master 2 slaves) Ubuntu Server 20.04

I followed the instructions from nekodaemon.com (I can't access the website right now) in the chapter "Slurm Quick installation for cluster on Ubuntu 20.04", but I removed the last line they say to add on the compute node

CgroupMountpoint=/sys/fs/cgroup because it created an error when launching the start

Process: 46877 ExecStart=/usr/sbin/slurmd $SLURMD_OPTIONS (code=exited, status=1/FAILURE)

May 02 10:15:54 ben1 systemd[1]: Starting Slurm node daemon...
May 02 10:15:54 ben1 slurmd[46877]: error: _parse_next_key: Parsing error at unrecognized key: CgroupMountpoint
May 02 10:15:54 ben1 slurmd[46877]: error: Parse error in file /etc/slurm-llnl/slurm.conf line 149: "CgroupMountpoint=/sys/fs/cgroup"
May 02 10:15:54 ben1 slurmd[46877]: fatal: Unable to process configuration file
May 02 10:15:54 ben1 systemd[1]: slurmd.service: Control process exited, code=exited, status=1/FAILURE
May 02 10:15:54 ben1 systemd[1]: slurmd.service: Failed with result 'exit-code'.
May 02 10:15:54 ben1 systemd[1]: Failed to start Slurm node daemon.

After this I was able to start munge and slurm on the master node but on the compute node:

I run:

sudo systemctl start slurmd

I get:

Job for slurmd.service failed because the control process exited with error code.
See "systemctl status slurmd.service" and "journalctl -xe" for details.

Then I run journalctl -xe and I get:

The job identifier is 22481 and the job result is failed.
May 02 10:48:48 ben1 sudo[47959]: pam_unix(sudo:session): session closed for user root
May 02 10:49:04 ben1 multipath[47985]: sdc: can't store path info
May 02 10:49:04 ben1 multipathd[771]: sdc: spurious uevent, path not found
May 02 10:49:04 ben1 multipathd[771]: uevent trigger error
May 02 10:49:05 ben1 multipath[47992]: sdc: can't store path info
May 02 10:49:06 ben1 multipathd[771]: sdc: spurious uevent, path not found
May 02 10:49:06 ben1 multipathd[771]: uevent trigger error
113
0 + 0
ธงชาติไทย
Elon Musk
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.