Affected services:

  • Compute nodes

Reduced capacity Peregrine V100 GPU nodes

Scheduled for Monday 16th March 2020 at 08:00 (Amsterdam)

Schedule/description of work

Due to maintenance and upgrades in the cloud infrastructure that hosts the Peregrine V100 GPU nodes, half of these nodes will be offline on March 16 and 17. All other Peregrine nodes, including the K40 GPU nodes, will be unaffected and will stay available. As a reminder, the following page on the wiki describes how you can select a specific type of GPU: https://wiki.hpc.rug.nl/peregrine/jobs/advanced_topics/running_jobs_on_gpus

Scheduled start time
March 16, 2020 08:00
Duration
2 days 5 hours
Status
Finished

Updates

We need a bit more time to complete the maintenance, and expect to bring the nodes back online tomorrow (March 18).

Posted by Bob