Completed: Quarterly Cluster Maintenance (July 16)

Dear Users,

We have completed our scheduled quarterly maintenance.  The MPI, SMP, GPU, HTC, and Invest  clusters have been returned to production.  There are a few residual problems we will take up in the morning.  Particularly:

  • is currently down
  • has network connectivity issues
  • some of the cluster nodes are in the down state

During this maintenance, we

  • improved networking
  • applied patches to interactive servers and compute nodes
  • applied a fix to issues with the /zfs1 filesystem
  • retired the compbio partition in the MPI cluster
  • retired the legacy partition in the MPI cluster
  • retire the titan partition in the GPU cluster
  • retired mobydisk

The aforementioned hardware were retired because they have reach end of life ( greater than 5 years old).  As always, questions and issues can be submitted to our ticketing system,


Thanks for your patience.  --CRC Team