Atlas: diagnose and fix disk pressure

Disk pressure

Symptom: Disk Usage metric near 100%, or writes failing with out-of-space errors. On the free tier the cap is small and this arrives fast.

Diagnose

1. Atlas Metrics: Disk Usage trend. Sudden jump suggests a bulk load or runaway job; steady climb suggests organic growth.
2. Check the biggest collections and indexes (Data Explorer storage sizes). Indexes can rival data size on write-heavy collections with many secondary indexes.
3. Check for unbounded growth: logs, events, or history collections with no TTL or archival.

Confirm

Identify the top storage consumers. If one collection dominates and it is append-only history, the fix is lifecycle management, not a bigger disk.

Fix, in order

• Drop unused collections and indexes (the Performance Advisor flags unused indexes).
• Add TTL indexes or an Online Archive rule for time-series-ish data.
• Compact only if fragmentation is proven; it needs the right role and takes time.
• Scale the tier for genuine growth, and set a disk alert at 75% so this never surprises you again.

Verify

Disk Usage drops and stays down for a full growth cycle. The alert is configured and tested.

Disk usage monitoring per https://www.mongodb.com/docs/atlas/monitor-cluster-metrics/; alert thresholds per https://www.mongodb.com/docs/atlas/alerts/; Online Archive for data lifecycle per https://www.mongodb.com/docs/atlas/online-archive/manage-online-archive/, verified 2026-09-26.