OIT Research Computing

Latest News

Coeus HPC cluster rebuild happening on October 28th 2024

Starting on the morning of Monday October 28th, the Coeus HPC cluster will be brought offline to complete our modernization project. This will include updating the operating system for all cluster nodes to a current supported version, updating environment modules, and updating our cluster management tools. We anticipate the cluster will be back online on Tuesday the 29th.

Given the large scope of updates, we will be canceling any running jobs at the start of the maintenance since it is unlikely the jobs will be able to complete if resumed. Once the cluster is available again, you may need to update your jobs. In particular, your module loads will need to be updated to use new versions of software. If you have any software that you have built in your home directory or other location, it may need to be rebuilt as well. As before, command `module avail` can be used to view the available modules. 

We understand that this will be disruptive to some users and we will be here to help! Please communicate with help-rc@pdx.edu with any questions, concerns, or issues that you may have.

What is changing?

• The cluster operating system will be upgraded from CentOS 7 (login, compute nodes) and Ubuntu 20.04 (GPU nodes) to Rocky 9
  
  CentOS 7 reached end-of-life this year and is no longer supported. Because the CentOS project no longer provides a long term stable release, we will be switching to the Rocky distribution which is compatible with CentOS and RedHat while providing long term support. The GPU nodes will also be switched to Rocky from Ubuntu to provide a consistent environment for users.
  
  

• Modules
  
  Due to the operating system upgrade, all of the cluster software modules must be updated and software rebuilt. We will not be recreating all of the existing modules in the new environment and instead will be creating a new set of core, widely used modules. Please let us know if you need a program that was not migrated to the new environment.
  
  New modules will include:

  • python/intel

  • cuda

  • gaussian

  • gcc

  • intel/mpi

  • mpich

  • openmpi

  • julia

  • r

  • rstudio

  • maven

What is not changing?

• Home directories, shares, and scratch space will not by modified

• Slurm scheduler and all commands (srun, sbatch, sinfo, squeue)

• Slurm partitions (short, medium, long, allcpu, gpu, priority access)

Thank you for your understanding, and we hope you will enjoy using the updated cluster environment!

Research Computing virtual Office Hours

OIT Research Computing team will host virtual Office hours on Monday December 18th 2023 from 2 pm to 2:50 pm. Feel free to stop by with any questions, issues, concerns or ideas that you would like to share with us. Starting at 2 pm, you will be able to join in using the following google meet link. 

Coeus maintenance and home directory change

We will be performing patching of the Coeus HPC on Friday, October 20th 2023 starting at 8:00 am. During the patching we will also remount the home directories to the different backend storage location and that will require us to reboot the login nodes and cancel any of the running jobs. Because of that, I strongly encourage you not to start any important long running jobs leading up to the patching date as they will have to be cancelled.

Home directories will be all migrated to our more robust enterprise storage array that will allow you to restore the files from the snapshots yourself by accessing the .snapshot directory located at the top of your home directory. Important change will be the enforcement of the 100GB user quota.

Research Computing virtual Office Hours

OIT Research Computing team will host virtual Office hours on Monday October 9th 2023 from 10 am to 11:30 am. Feel free to stop by with any questions, issues, concerns or ideas that you would like to share with us. Starting at 10 am, you will be able to join in using the following zoom link. 

Priority Access Now Available in OIT High Performance Computing (HPC) Cluster

It is now possible to reserve priority access in OIT’s Coeus HPC cluster. In addition to the Free access tier, there is now a Priority access tier making it possible for researchers to reserve dedicated computer time for their funded research needs. Details are available in OIT’s description of the High Performance Computing (HPC) Clusters service, including a link to the HPC Priority Access request form where researchers can engage with OIT to assess their HPC requirements in order to include funding for Priority access in future research grant proposals.

Nvidia V100 GP-GPU's to be added to the Coeus HPC 

The demand for general purpose Graphics Processing Units for high performance computing has been grown rapidly. OIT has purchased two Nvidia Tesla V100 GPU's. These are installed in the two Coeus himem nodes.  These each have one Tesla V100 GPU, 2 CPUs 24 cores, 768 GB RAM, and 1.9TB local scratch.

More information on how to run jobs on these nodes can be found here:  Using Coeus Himem GPUs

Campus CyberInfrastructure

The Office of Information Technologies - Research Computing (OIT-RC) supports an extensive infrastructure to facilitate academic research computing activities across the PSU campus.  OIT coordinates with campus research faculty, administration, and regional partners to offer the most extensive, cost-effective, and secure systems possible.  The OIT website has a good high-level overview of the full scope of the Research Computing CyberInfrastructure.  This includes information about data storage, web, GIS, and database serving maintained by OIT.  

Also available is a slide presentation on PSU Research Computing Infrastructure.

This site is focused specifically on supporting research computing on these systems.