HPC Status Page
Affected services:
Scheduled for Wednesday 12th October 2022 at 06:00 (Amsterdam)
Schedule/description of work
Unfortunately the power system maintenance, announced earlier, has been moved from October 5-7 to October 12-14. From Wednesday morning 12 October until Thursday morning 13 October maintenance will be performed on the power distribution systems at the DUO data center, where the Peregrine cluster is located. This will cause the Peregrine compute nodes to lose power a few times during this period. To prevent unexpected job crashes, we have scheduled a maintenance reservation in the system preventing jobs from running during the power outages. This means long running jobs will not start if these may not finish in time. We will make use of the time between outages to perform some file system checks, which can only be done while the file systems are not in use. Unfortunately there does not seem to be enough time between the outages to make the cluster available in the period in between them. We expect to bring the systems back online on Thursday morning.