We will be performing a second scheduled maintenance on Friday, November 6, beginning in the morning.
During this time, we will be decommissioning both the old /home/rcf-proj file system and the old HPC cluster’s main partition. This means that, after November 6, you will no longer be able to access any data still stored in /home/rcf-proj or run any jobs on the old HPC cluster main partition. In other words, you will only be able to use the new Discovery cluster for your jobs after November 6. If you need more information on using Discovery, please see our Getting Started with Discovery user guide.
During this maintenance period, all other systems (Discovery login nodes, /project, and the /scratch and /scratch2 file systems) will still be accessible, although you may experience some intermittent downtime. However, you will be unable to run/submit jobs during the maintenance period as we perform maintenance on the Discovery compute nodes. If a job is submitted before the maintenance but won’t have enough time to complete before maintenance begins, it will be rescheduled to start after the maintenance is completed.
Note: If you have condo nodes on the old HPC cluster, your nodes will not unexpectedly become unavailable after this maintenance. If you have not completed your data migration by November 6, we will continue to work with you to discuss condo node migration plans.