Fil des news

We need to perform a maintenance to E5-GPU partition's nodes. Nodes are in draining mode, jobs will not start until maintenance is done.

EDIT, 2024-12-05: Most nodes done, queue OK.

Following a night of heavy network problems, our slurmctl is DOWN. We are working on it.

EDIT 11:00: slurmctl is back ONLINE.

Following the power outage of this week-end, we are currently still upgrading our servers.

Follow this newsfeed to stay updated.

EDIT 14:20: We are restarting. Still, not all nodes are available.

EDIT 16:30: Full restart.

kind reminder: a complete power outage is scheduled Saturday Oct 26th 2024 (tomorrow).

as of now, we are starting to shutdow our HPC infrastructure (partitions, then gateways, then nodes and filers).

• partitions are DOWN
• gateways are OFF
• upgrades on nodes and filers (IN PROGRESS)

An electrical power outage for maintenance purpose will be held Saturday 26th of October 2024 (10 hours long maintenance).

• All PSMN infrastructure will be powered off from friday 25th (starting 12h00) to tuesday 28th
  • login nodes, compute nodes & visualization nodes will be POWERED OFF at noon (12h00) friday

• At restart:
  • “flix” partitions will be renamed “premium”, watch your scripts.
  • E5 and Epyc hardware might be moved within DC, hence not available at restart (depending on humans availability)

See our online documentation:

• https://www.ens-lyon.fr/PSMN/Documentation/