🔒 Paid
Missing Database Index
A recently deployed order-history query path is performing full-table scans as traffic to it rises. Per-account order history must remain available to customers, and the Orders Service's connection pool is shared with other order-processing traffic — exhausting it risks degrading unrelated checkout paths too. Diagnose why database CPU and query latency are climbing together and bring the primary back to a durable, sustainable state.
- Difficulty
- Intermediate
- Expected Time
- ~12 min
- Category
- Databases & Storage
Skills
- Diagnosing full-table-scan query regressions
- Distinguishing database-side contention from API-layer capacity
- Recognizing when scaling the wrong tier makes contention worse
- Evaluating feature-sacrifice mitigations versus root-cause fixes
System Architecture
- Clients
- API Service
- Orders Service
- PostgreSQL Primary
This is a Pro scenario. Upgrade to Pro to unlock it.
Upgrade to ProBack to Incidents