Fixed
All Worker Nodes are back online. Downtime is officially finished. You can start sending jobs again.
Watching
All Worker Nodes have been stopped.
Because of self-cleaning, /user might also be slow during the weekend.
From the Technical Team, it seems operations will still happen on Monday, after which hopefully cluster will be restarted.
Reported
Dear Users,
The timeline for the downtime has been finalized by the team working on the building.
The downtime will officially be from Friday 17/11 4PM to Monday 20/11 8PM. The compute nodes will all be stopped Friday around 4PM. Jobs will stop running, and we expect them to be rescheduled and queued.
During the downtime, all storage (/pnfs, /user) will stay accessible, as well as mX machines and services. Batch system will receive jobs, but none will be able to start until compute nodes are started again.
While the downtime extends up to Monday evening, it is possible, if everything goes well, that work will be finished on Sunday. So compute nodes might be started early, please follow status.iihe.ac.be for follow-ups.
Please note that as the works include touching at the electric lines delivering power to the Datacenter, while failover equipments and procedures are implemented, there is a small risk that all power to the cluster will be cut. In that case, all storage and services would come down.
Cheers, The T2B IT Team