Skip to main content
Have a request for an upcoming news/science story? Submit a Request

Snyder cluster retirement

  • Outages and Maintenance
  • Snyder

Link to update at January 13, 2021 11:07am EST UPDATE:

To give research teams more time for data migration, we are pushing Snyder A and B nodes retirement date to January 29th, 2021. The transition plan for the C, D, E, F nodes remains the same.

Link to original posting ORIGINAL:

The bulk of the Snyder cluster (A and B nodes of 2015 vintage) has reached the end of its life cycle and is being retired on Friday, January 15th, 2021. Researchers owning nodes in A or B sub-clusters should start archiving any data they may have there to the Fortress Archive now, or move it to other clusters or Data Depot for any ongoing work. Please note that after the cluster is retired, you will not be able to retrieve any data left in the Snyder scratch filesystem.

Shares in our latest community cluster, Bell, are available for purchase. Bell will have considerable advantages over Snyder in many respects -- the nodes contain many more processing cores that can be scheduled for jobs, and will provide more ability to take advantage of standby scheduling. Each Bell node contains 256 GB of RAM. In addition, Bell has faster networking, and a newer technology for its scratch filesystem that will improve your processing efficiency.

A small handful of Snyder nodes (the C, D, E and F subclusters of 2016 and 2018 vintage) still have remaining life time in them. Partners of these resources have been contacted individually to discuss the plan for moving their allocations to Bell.

Please contact us at rcac-help@purdue.edu if you have any questions.

Originally posted:
Last updated: