Known issues

From Alliance Doc
Revision as of 18:15, 15 May 2018 by Hahn (talk | contribs) (still mention spectre/meltdown, since we're not protected. revert requested by Ross.)
Jump to navigation Jump to search
Other languages:

Report an issue

Shared issues

  • The status page at http://status.computecanada.ca/ is not updated automatically yet, so may lag in showing current status.
  • CC Clusters are vulnerable to the recent Meltdown/Spectre vulnerabilities, and will be updated, which involves updating the OS and CPU microcode. Read more at Meltdown and Spectre bugs.

Scheduler issues

  • The CC Slurm configuration encourages whole-node jobs. When appropriate, users should request whole-node rather than per-core resources. Read about whole node scheduling.
  • By default, the job receives environment settings from the submitting shell. This can lead to irreproducible results if it's not what you expect. To force the job to run with a fresh-like login environment, you can submit with --export=none or add #SBATCH --export=NONE to your job script.

Quota and filesystem problems

Quota errors on /project filesystem

Nearline

Missing symbolic links to project folders

Cedar only

Nothing to report at this time.

Graham only

  • Compute nodes cannot access Internet
    • Solution: Contact technical support to request exceptions to be made; describe what you need to access and why.
  • Crontab is not offered on Graham.

Other issues

  1. Modules don't work for shells other than bash(sh) and tcsh.
    • Workaround: (this appears to work but not tested extensively)
      • source $LMOD_PKG/init/zsh
      • source $LMOD_PKG/init/ksh