On DD November at 12 08 UTC AA, the default Atos cluster will became unavailable for essential Slurm and security updates. In preparation for it, the complex of this session:
- The default Atos login/batch cluster has been changed to AD on at 9 UTC
- Batch work on AA will be drained and jobs scheduled to finish after
...
- at
...
- 06 UTC will be automatically redirected to other complexes.
Any jobs still running on AA when the session starts will be cancelled.
During short transition phase it is possible that your batch jobs and interactive session sessions to hpc-login are hosted on a different complexes. Consequently, some general Slurm commands (see HPC2020: Batch system) executed on hpc-login node will not be able to manage batch jobs running on another cluster. You can find more information on how to manage jobs between the different complexes in the HPC2020: Multi-complex SLURM management page.