UPDATE 8/25: Campus electricians have delayed full return to service until 8/31; we'll have limited power before that to enable maintenance and storage bringup.
UPDATE 8/23: One of the old transformers is powered-off and work is progressing.
Starting Tuesday, August 21, 2018 at 7 a.m. ET, the Graham cluster will be unavailable to all users and running jobs will be terminated. The power feed to the building is being upgraded, which requires a complete shutdown. This is a major project that will take about eight (8) days to complete. The Graham cluster, all storage, Globus, cloud, TSM backup services and all associated hardware will be unavailable to users during the outage.
Users are encouraged to migrate to Cedar before the outage and to shutdown their running virtual machines in Graham Cloud before the downtime. If you have used Niagara in the past, you can also consider this cluster as an alternative.
Please watch http://status.computecanada.ca for updates on the availability of Graham and all other national systems.
Start Date : Tuesday, August 21, 2018 Start Time : 7:00 a.m. ET Anticipated End Date : Wednesday, August 29, 2018
Users will be notified by email when the cluster is up and running again.
The duration of this outage is subject to change; updates will be posted to http://status.computecanada.ca and via the Compute Canada, Compute Ontario, and SHARCNET Twitter accounts. If the outage is extended by more than two days, another user communication will be sent directly.
For questions, or assistance migrating to other national systems, please email firstname.lastname@example.org