CHTC outages starting 5pm today (3/9)

[chtc-users@xxxxxxxxxxx]

Greetings CHTC Users,

 

Due to short-notice maintenance for the cooling infrastructure in one of CHTCâs server rooms, we will have an outage starting tonight (3/9) that will impact our HPC cluster and certain HTC execute nodes. 

 

HPC Cluster

The HPC cluster, including execute nodes, head nodes, and filesystem, will be down starting at 6:00pm on Wednesday (3/9) with services expected to be restored by the end of the day on Thursday (3/10). 

The HPC cluster is configured to not start new jobs that will not complete before the start of the scheduled downtime. Already-running jobs that do not complete by tonight will be interrupted and re-queued once systems are operational. Users may want to remove any jobs that will be negatively impacted by this process. 

 

HTC System

Some HTC execute nodes, including some researcher-owned GPU nodes, will be down starting at 5:00pm on Wednesday (3/9) with services expected to be restored by the end of the day on Thursday (3/10). Jobs running on these nodes will be evicted and remain in the queue to be re-run. 

We do not expect the HTC system submit servers, transfer server, or Staging filesystem to be impacted at this time. 

 

We appreciate your patience as the ongoing maintenance in this room has meant a lot of extra announcements to this list, and not always with our usual amount of notice.

 

As always, email us at chtc@xxxxxxxxxxx with any questions. 

 

Regards,

Your CHTC Team

 

 

-- 

Christina Koch (she/her)

Research Computing Facilitator

Center for High Throughput Computing - University of Wisconsin - Madison

OSG and OSG Connect

email: ckoch5@xxxxxxxx // calendar: https://go.wisc.edu/clk-calendar