The majority of HPC resources will remain down through the afternoon of Monday, June 9. A disk enclosure subsystem of the Derecho scratch filesystem has failed, requiring further replacement activities. A spare part is being shipped, but we will not be able to restore additional services today.
The Campaign Storage, Home, and Work file systems are up and accessible through the Casper login nodes and Globus. The Quasar tape archive and Stratus object storage systems are also available.
All Derecho components, Casper compute nodes, and JupyterHub will remain offline until the scratch filesystem is restored.
On Monday morning, we will replace the affected part and regain access to the scratch filesystem. This will allow us to begin restoring additional Casper resources, JupyterHub, and Derecho. The Monday timeline will depend on the progress of the filesystem repair, but we intend to have Casper - and possibly Derecho - available by the end of the day, with all resources returned by Tuesday morning, June 10 if additional time should be needed.
Maintenance activities for NWSC facility mission-critical infrastructure are now complete. This work included major upgrades to electrical and mechanical subsystems throughout the facility, replacing aging components and performing additional preventative maintenance.