- Kraftig Linux tungrekneanlegg som er designa for å handtere mange samtidige jobbar og brukarar med varierende behov.
Kven tar eg kontakt med?
Treng du hjelp med tenestene eller råd om korleis du kan utnytte rekneressursane best mogleg?
- Send epost til firstname.lastname@example.org.
[SOLVED] Upgrade of Queue system
2. nov. 2017 10:09
We will upgrade the queue system on Abel in a few minutes. During the upgrade, jobs will run as normal, but the queue system commands will be unavailable. We expect the upgrade to take about 15 minutes.
Update: The upgrade is now done. We don't expect any significant user visible changes.
Error in the queue system
16. okt. 2017 12:37
Slurm was upgraded on Abel last Thursday evening. It was supposed to be a non-noticeable upgrade, but unfortunately the rpms we installed had an error. This led to error messages in slurm-NNN.out like
/cluster/bin/setup.sh: line 72: /work/jobs/18612887.d/.tmp/machines: No such file or directory
/var/spool/slurmd/job18612887/slurm_script: line 37: sstat: command not found
on some nodes. Many jobs will have run fine anyway, but some (especially the ones needing the "machines" file) will have crashed. The problem has been fixed now, and we are sorry for the inconvenience.