trixie, slurmctls problem, troubleshooting loop
 

Dear Debian experts:

System
testing (trixie) amd64 A while ago, this penguin crashed after dist-upgrade, then resuscitated (minimal system clean installed, many packages installed following apt log)

Problem
Cluster system (slurm) stopped working.

Troubleshooting
1. Following logs, it looks like that slurmctld fails to kick in.
2. systemctl status slurmctld.service says that "slurmctld: fatal: Can't find plugin for select/cons_res"
3. For some reason, /usr/lib/x86_64-linux-gnu/slurm-wlm/select_cons_res.so is missing in current repo, while bookworm repo has it.

4. To rule out some mix-up, I tried to purge slurmctld, and install.
aptitude purge slurmctld says that pre-removal script fails.
5. Followed /var/lib/dpkg/info/slurmctls.prerm, this script complains that "slurm_load_jobs error: Unable to contact slurm controller (connect failure)"
6. goto 1. above.

Question
Am I on the right track?
Is there any good idea to get out of this loop?
Another system upgrade does not help.

Any information will be appreciated

I have just tried to install trixie amd64, did not work. I guess it is still incomplete.

Hya

pan64, Thank you for your time.

cheers

The t64 transition.
 

Hi.

Debian is currently undergoing a huge transition, replacing 32-bit time_t with 64-bit time_t.

It's not recommended to force upgrades or do fresh installs (that is for Trixie/Sid) till things settle down.

https://wiki.debian.org/ReleaseGoals...al_description