Thursday, June 5, 2014 - 3:34pm
The High Performance Computing team is pleased to announce the upcoming availability of a new HPC cluster Grace. Final acceptance testing of the cluster is in progress and the process of decommissioning the Bulldog-J and Bulldog-K clusters has begun. Over the next (2) months, data and jobs will need to be migrated to other clusters.
What do I need to know?
The schedule of events to support the decommissioning is as follows:
- May 27: Job submission to the fas_very_long queue will be suspended.
- June 16: Job submission to all other queues will be suspended.
- June 27: All compute nodes will be turned off by this date; any active jobs will be cancelled.
- July 11: All data must be moved off BulldogJ and BulldogK by this date.
What should you do now?
- Plan ahead for the suspension of the job queues on the dates listed.
- Begin identifying data that needs to be retained and moved elsewhere. It is strongly recommended that unnecessary data be deleted at this time prior to migration. It should be expected that 1Tb of data will take approximately 10 hours to be transferred to a new location, however, individual results depend on network traffic in and out of the clusters, as well as your local network connection.
Assistance will be available any researchers who need help to identify a new location for their data and jobs, or help with moving the data. If you have any questions or concerns about this work or the schedule, please contact hpc@yale.edu .
Thank you in advance for your cooperation.