Unscheduled power outage in SRCF
All services have been restored and Sherlock is running normally again.
Power has been restored at the data center, we’re currently working on bringing the affected parts of the cluster back up.
We’re currently investigating a power outage affecting a significant portion of SRCF, the datacenter where Sherlock is hosted. All the Sherlock 4.0 compute and service nodes are currently down, which is impacting job wait times, availability of owner partitions and potentially network connectivity to the cluster. Engineers are en route to investigate and hopefully restore power as quickly as possible. Updates to follow.
4 Affected Services:
- Login / login.sherlock.stanford.edu
- Job scheduler / Slurm database
- Job scheduler / Public partitions
- Job scheduler / Owner partitions