In order to allow the university to perform a generator test on the campus data center, the 28-core, Tesla k80 gpu, Tesla p100, and Tesla v100 queues and login nodes login1 and login2 will be going offline for scheduled maintenance at 4pm on Monday, March 11th. The maintenance is expected to conclude by lunch time on Tuesday, March 12th.
We thank you for your patience while these necessary tests are conducted.
Four of our Intel Sapphire Rapids nodes have been updated to include 1 TB of DDR5 memory each. In order to simplify the user experience, the high bandwidth memory on these nodes has also been reconfigured from main memory to level 4 cache. These nodes are now accessible via the hbm-1tb-long-96core queue. For more information, please see our full list of SeaWulf queues.
UPDATE: the authentication issue described below has been resolved, and new connections to SeaWulf are no longer failing.
The campus Identity server, which we rely on for authenticating access to SeaWulf, is currently experiencing issues, resulting in new connections to our HPC environment failing.
We will be performing upgrades & maintenance on the SeaWulf Storage on Monday October 9th starting at 9:00 AM. During this maintenance window, all Seawulf login nodes and queues, as well as the storage, will NOT be available. The SeaWulf cluster is scheduled to return to normal operation by the end of business on Tuesday October 10th.
We thank you for your patience while these necessary upgrades are completed.
We have just been notified of scheduled electrical maintenance at the Campus Data Center on Tuesday August 22nd. During these necessary electrical upgrades the 28-core and GPU (K80) queues will be off-line and jobs running on those queues will be terminated, starting at 9:00 AM on Tuesday the 22nd. The maintenance is expected to be completed by the end of business on the 22nd.
No other queues will be affected, and the login nodes will remain accessible during this maintenance period.
We thank you for your patience while these necessary upgrades are completed.
In preparation for a new expansion to SeaWulf (announcement to follow), we will be performing maintenance on the 96-core, 40-core, and A100-GPU queues, as well as the milan.seawulf.stonybrook.edu login nodes, starting at 10:00 AM on Wednesday June 14th and concluding by the end of business the same day.
During this maintenance window, the 28-core and GPU queues, as well as the login.seawulf.stonybrook.edu login nodes will continue to be available.
We thank you for your patience while we perform these necessary preparations.
In order to allow the university to perform electrical maintenance on the circuits for the 24-core queues and the login node 'Login2', they will be offline starting Sunday evening April 16th concluding Tuesday April 18th by the end of the day. No other queues will be affected, and the login node 'Login1' will remain accessible during this maintenance period.
We thank you for your patience while these necessary upgrades are implemented.
We have just been notified by the manufacturer of a critical firmware update that needs to be applied to SeaWulf to ensure we can continue to provide a robust computational environment. Out of an abundance of caution, we are taking an emergency maintenance window, scheduled for ~3 hours from now at 3:30 PM today. This window is expected to last for ~24 hours.
During this maintenance window, all Seawulf login nodes and queues, as well as the storage will NOT be available.
We apologize for this unforeseen development, and thank you for your patience while we apply these critical firmware updates.
In order to allow the university to perform electrical maintenance on the campus data center, the 28-core and Tesla K80 gpu queues will be going offline starting at end-of-business on March 21st. These queues will come back online after the maintenance is concluded by the end of business on March 22nd. No other queues will be affected, and the login nodes will remain accessible during this maintenance period.
We thank you for your patience while these necessary upgrades are implemented.
In order to perform the 2nd phase of updates on the SeaWulf storage arrays, upgrade the networking and perform electrical maintenance on the infrastructure, the SeaWulf cluster will be going off-line starting 9:00 AM on Wednesday February 15th. The maintenance is expected to be completed by the end of business on Thursday February 16th. During this maintenance window, all login nodes, compute nodes as well as the storage array will NOT be accessible.
We thank you for your patience while we perform these necessary updates.
