Affected deployments were moved to other agents, while existing containers are running fine. We're working on restoring the affected agents to accept new deployments.
Update 1: The defective agent was restarted, remaining containers from that host are running again. We're monitoring.
Update 2: Everything is up and running.