Cluster Maintenance Important Changes

Hello All,

Thank you for your patience during this downtime. It was a long one, but we are really excited about the changes. The HTC cluster is now part of the H2P supercluster! This update will greatly simplify management of the cluster and allow us to forge a path forward for new hardware. However, there is likely to be some growing pains with such a large update (mainly HTC users). Please continue to use to submit help tickets. Below, I will highlight the important changes:

Updates to H2P (smp, mpi, gpu, htc, and Invest, where applicable)

Update to Red Hat Enterprise Linux 7.6

Update to OmniPath 10.3

Update to CUDA 10.0.130 w/ NVIDIA driver 410.79

Update to Slurm 17.11.7

Updates to HTC

New login node: (note change from sam to crc)

Groups are now charged service units for usage

10,000 given for free

Details for proposals up to 500K can be found at

Use `` on the command line to see your limit and are using the new queuing system

OpenOndemand has a new hostname:

Outstanding Issues will be down until it can be upgraded to RHEL 7 needs a new SSL certificate



Happy Computing,

The CRC Team