During a routine cloud-agents control plane release, a configuration error caused the service to crash. We quickly identified and corrected the problematic config value, restoring normal operation. We are investigating ways to make our configs more robust and will look to add additional testing around this process to ensure the issue doesn’t happen again.