RCF Decommission

RCF Decommission:

• What is happening: As per earlier announcement(s), the RCF HPC and tool servers are being permanently decommissioned. 1) Due to their age, those machines are becoming increasingly unreliable, and 2) This will help us focus our development and maintenance efforts on our newer resources – particularly our 132 node partition in Sherlock.\n

• What has already happened: Access to RCF has been shut off for a few weeks and we are performing a final backup of the filesystem.
• Systems affected: CEES-RCF HPC and tool servers (tools <7).
• Systems NOT affected: CEES-Mazama HPC and tools (tools >7 – noting that tool-7 is down, in an unrelated event), or Sherlock systems.
• When: Soon. The tentative timeline is mid-August.
• What you need to do: Probably nothing. If you do not use the RCF HPC or tool nodes (tools <7), this does not affect you. If you believe you data on RCF that needs to be recovered, let us know and we will make the necessary arrangements. Note that, as stated above, access to RCF has been shut down for weeks. If you didn’t notice that, you probably don’t use RCF :sunglasses:.

If you have questions or comments, contact Randy White or Mark Yoder via the CEES Slack channel.