At 9:31 PM PST, cluster maintenance on the Dragon cluster resulted in unexpected downtime and all agents being unavailable.

The incident was mitigated by Contextual on-call engineers and all agents returned to proper function at 10:35 PM PST.