The purchasing contract for the next generation of Argon HPC hardware has been finalized. We are planning to place an initial order in late September/early October for new infrastructure, replacement/expansion of UI queues, and any initial investment desired by the HPC community.

Please join us in welcoming Josh Johnson to Research Services!

Please join us in welcoming Gabby Perez back to Research Services!

To replace storage systems that provide user homes in Argon, we will be deploying a storage cluster using the Ceph distributed filesystem. This cluster is designed to improve performance and reduce latency for data access in cluster users' home directories as well as reduce the administrative burden for Research Services staff. Future growth and hardware replacement will be significantly simplified by using a storage cluster...

Update (03-16-2022, 8:15 AM): In order to fully resolve this issue, compute nodes need to be rebooted. Reboot jobs will be launched on the cluster to opportunistically reboot nodes. No user jobs will be removed but there may be some delays in job launches as some number of nodes may be offline at any given time. Argon HPC is experiencing an issue with file locking on network shares.  This may appear as frozen or hung applications, or, jobs failing with errors indicating an inability to create lock files...