database-observability

What problem does it solve?

Database performance is often opaque at the query level, making it difficult to find slow queries, understand root causes, and correlate database behavior with application traces; this Skill surfaces query samples, RED metrics, and visual explain plans to pinpoint and diagnose performance issues.

Core Features & Use Cases

• Query performance dashboard: Auto-provisioned views for top queries by total time, call count, and mean latency.
• Query samples & explain plans: Capture actual query parameters, timing, and visual explain plans to identify index usage and scan types.
• RED metrics & alerts: Per-query RED metrics, Prometheus alert rules for slow queries, high error rates, and connection saturation.
• Trace correlation: Link db.statement and OTel attributes to application traces to trace slow queries back to services and requests.
• Deployment scenarios: Works with PostgreSQL and MySQL across self-managed and cloud variants (RDS, Aurora, Cloud SQL, Azure).

Quick Start

Enable pg_stat_statements or MySQL Performance Schema, create a least-privilege monitoring user, add the appropriate database_observability.* block to your Alloy configuration to forward metrics and logs, and verify the Database Observability dashboard appears in Grafana Cloud.