Event Date | Summary |
---|---|
A login node is available on Graham now, mainly to provide access to files. The migration of /project is almost complete, and when it is, we'll enable read-write access to /project. Until then, /home and /scratch filesystems are normal, but you won't be able to change your files or directories on /project. A limited set of existing Graham compute nodes are being configured, and will be made available to Slurm for jobs this week. |
|
Due to cooling and electrical issues in the datacenter, we are delayed in returning Graham access. We expect to open access to Graham logins on or before the 16th this will include /home and /scratch filesystems, but the transfer of /project (to new hardware) is not complete. Also, compute nodes (via Slurm) will be gradually returned to service, with a simplified set of partitions. |
|
The power and cooling outages over the holiday break went well but one required power component is still not back in service. We will not be able to bring back any graham compute resources until we have that power feed available. The current ETA is Jan 13th |
Service | Incident status | Start Date | End Date |
---|---|---|---|
Graham | Closed |
2024-12-07 - Planned Outage - Arrêt planifié
Renovations for the upgrade require a full datacentre shutdown from Dec 7th to Jan 16rd.
During this time no services at the Graham datacentre will be available: this includes all cluster, storage, cloud, and related services.
===
Graham ne sera pas disponible du 7 décembre au 16 janvier.
Pendant cette période, aucun service au centre de données de Graham ne sera disponible. Cela inclut toute la grappe de calcul, l'accès au stockage, l'infonuagique et les services reliés.
Updated by Mark Hahn on