Network issue 2025 03 31

A brief network outage on Monday evening, March 31, 2025, caused most Euler VI compute nodes and Nvidia RTX 3090 GPU nodes to go offline.

All jobs running on these nodes failed and their state will be shown as "NODE_FAIL" in Slurm. Please resubmit them.

Sorry for the inconvenience.

Edit: a first version of this article defined this incident wrongly as a power incident at CSCS