Lessons from a full disk: capacity monitoring needs ownership

A full disk is rarely a sudden event. Logs, temporary files, retained installers, databases, and backups usually provide warning through steady growth. The outage happens when no one owns the threshold or the response. Use capacity monitoring that shows both current utilization and growth rate. Define warning and critical thresholds that leave enough time for action, and send alerts to a team that can investigate. Pair every alert with a runbook: which paths to inspect, what may be cleaned safely, and when storage expansion or retention changes require approval. After a capacity incident, retain the lesson in the service record. Review the trend, correct the immediate pressure, and fix the missing control that allowed it to become an outage.