The rollback has completed and errors have subsided.

An incorrect configuration change applies aggressive throttling on container filesystem bandwidth and is causing downtime.

We have reverted the configuration and are managing recovery.