High Performance Computing Service

Relocation Schedule

Last update: Tue Aug 13 00:32:53 BST 2013

The HPCS cluster will be moving to a new modern machine room in July 2013. The current plan is outlined on this page and any amendments will be reflected here.

When  Event  Comments
  
[DONE] Tuesday 23rd July 12:00  Temporary reduction in capacity  Withdrawal of all Westmeres and 25% of Teslas, and reduction of the number of available Sandy Bridge nodes by 40%. Filesystems and login access, and 60% of the Sandy Bridge compute nodes, will remain available for normal use until 30th July.
  
[DONE] Thursday 25th July  Phase 1 move  Disconnected nodes transferred to the new room.
  
[DONE] Tuesday 30th July 08:00  Service suspended  No service - total blackout including filesystems. Preparations for the move of the remaining equipment.
  
[DONE] Wednesday 31st July  Phase 2 move  Remaining hardware moved to the new room, followed by recommissioning work.
  
[DONE] Friday 2nd August  Racking and cabling completed.  
  
[DONE] Saturday 3rd August  Login and filesystem access restored  
  
Tuesday 6th August 03:13  Teslas and 50% Sandy Bridge compute nodes restored to service.  
  
Friday 9th August 22:00  Westmeres and more Sandy Bridge compute nodes restored to service.  
  
Tuesday 13th August 00:00  Remaining Sandy Bridge compute nodes restored to service. Service enters normal operation.