Advanced PostgreSQL query performance monitoring, SQL optimization, and database analytics extension
Monitor slow queries • Track cache efficiency • Analyze WAL generation • Optimize database performance • Real-time metrics • Grafana dashboards
English (current) | 简体中文 | 繁體中文 | 日本語
Track 52 Metrics Across 11 Views - Monitor PostgreSQL Query Performance in Real-Time
Production-ready extension for PostgreSQL 16, 17, 18 - Drop-in replacement for pg_stat_statements with enhanced analytics
PostgreSQL Query Performance Monitoring Made Simple
pg_stat_insights is an advanced PostgreSQL extension for database performance monitoring, query optimization, and SQL analytics. Track and analyze 52 comprehensive metrics across 11 pre-built views to identify slow queries, optimize cache performance, and monitor database health in real-time.
Perfect for:
- Database Administrators monitoring PostgreSQL performance
- DevOps teams tracking query performance and resource usage
- Developers optimizing SQL queries and database operations
- SREs implementing database monitoring and alerting
Key Features:
- 52 metric columns - Execution time, cache hits, WAL generation, JIT stats, buffer I/O
- 11 pre-built views - Instant access to top slow queries, cache misses, I/O intensive operations
- 11 parameters - Fine-tune tracking, histograms, and statistics collection
- Drop-in replacement for pg_stat_statements with enhanced metrics
- PostgreSQL 16-18 - Full compatibility with PostgreSQL 16, 17, and 18
- 22 regression tests - Comprehensive test coverage for all features
- Response time tracking - Categorize queries by execution time (less than 1ms to greater than 10s)
- Cache analysis - Identify buffer cache inefficiencies and optimization opportunities
- WAL monitoring - Track write-ahead log generation per query
- Advanced features - JSON/JSONB, arrays, partitioning, triggers, window functions
- Time-series data - Historical performance trending and bucket analysis
- Prometheus/Grafana ready - Pre-built dashboards and alert rules included
- CI/CD ready - GitHub Actions workflows for multi-version testing
Monitor PostgreSQL query performance in under 5 minutes:
-- Step 1: Enable extension in PostgreSQL configuration
ALTER SYSTEM SET shared_preload_libraries = 'pg_stat_insights';
-- Restart PostgreSQL server required
-- Step 2: Create the extension in your database
CREATE EXTENSION pg_stat_insights;
-- Step 3: View your slowest queries instantly
SELECT
query,
calls,
total_exec_time,
mean_exec_time,
rows
FROM pg_stat_insights_top_by_time
LIMIT 10;Result: Instant visibility into query performance, execution times, cache efficiency, and resource usage across your PostgreSQL database.
Complete documentation available at:
Quick Links:
- Getting Started - Installation and setup
- Configuration - All 11 parameters
- Views Reference - All 11 views
- Metrics Guide - All 52 columns
- Usage Examples - 50+ SQL queries
- Prometheus & Grafana - Monitoring integration
- Troubleshooting - Common issues
# Build and install
cd pg_stat_insights
make
sudo make install
# Configure
echo "shared_preload_libraries = 'pg_stat_insights'" | \
sudo tee -a /etc/postgresql/*/main/postgresql.conf
# Restart PostgreSQL
sudo systemctl restart postgresql
# Create extension
psql -d your_database -c "CREATE EXTENSION pg_stat_insights;"Detailed instructions: Installation Guide
All 11 pre-built views:
| View | Purpose |
|---|---|
pg_stat_insights |
Main statistics view (52 columns) |
pg_stat_insights_top_by_time |
Slowest queries by total time |
pg_stat_insights_top_by_calls |
Most frequently called queries |
pg_stat_insights_top_by_io |
Highest I/O consumers |
pg_stat_insights_top_cache_misses |
Poor cache performers |
pg_stat_insights_slow_queries |
Queries with mean time > 100ms |
pg_stat_insights_errors |
Queries with errors |
pg_stat_insights_plan_errors |
Plan estimation issues |
pg_stat_insights_histogram_summary |
Response time distribution |
pg_stat_insights_by_bucket |
Time-series aggregation |
pg_stat_insights_replication |
Replication monitoring |
Complete reference: Views Documentation
Solve Common PostgreSQL Performance Problems:
- Find slow queries - Identify queries consuming excessive execution time and resources
- Optimize cache usage - Detect buffer cache misses and improve shared_buffers efficiency
- Reduce WAL overhead - Monitor write-ahead log generation per query type
- Track query patterns - Analyze execution frequency, response times, and resource consumption
- Monitor in real-time - Integrate with Grafana for live dashboards and alerting
- PostgreSQL best practices - Built following PostgreSQL coding standards and conventions
| Feature | pg_stat_statements | pg_stat_monitor | pg_stat_insights |
|---|---|---|---|
| Metric Columns | 44 | 58 | 52 |
| Pre-built Views | 2 | 5 | 11 |
| Configuration Options | 5 | 12 | 11 |
| Cache Analysis | Basic | Basic | Enhanced with ratios |
| Response Time Categories | No | No | Yes (<1ms to >10s) |
| Time-Series Tracking | No | No | Yes (bucket-based) |
| TAP Test Coverage | Standard | Limited | 150 tests, 100% coverage |
| Documentation | Basic | Medium | 30+ pages on GitHub Pages |
| Prometheus Integration | Manual | Manual | Pre-built queries & dashboards |
See detailed comparison: Feature Comparison
Comprehensive TAP Test Suite for Quality Assurance:
- 16 test files covering all extension functionality
- 150 test cases with 100% code coverage
- Tests all 52 metric columns, 11 views, 11 parameters
- Custom StatsInsightManager.pm framework
- No external Perl dependencies required
- Compatible with PostgreSQL 18 testing infrastructure
Run PostgreSQL extension tests:
./run_all_tests.shContinuous Integration: GitHub Actions workflow for automated PostgreSQL testing on every commit
Learn more: Testing Guide
Real-Time PostgreSQL Metrics Visualization:
Turn pg_stat_insights data into actionable dashboards and alerts for PostgreSQL database monitoring:
- Prometheus integration - 5 pre-configured queries for postgres_exporter
- Grafana dashboards - 8 ready-to-use panels for query performance visualization
- Alert rules - 11 production-ready alerts for database health monitoring
- Query rate tracking - Monitor queries per second (QPS) and throughput
- Cache performance - Real-time buffer cache hit ratio monitoring
- Response time SLA - Track P95/P99 query latency for service level objectives
- WAL generation alerts - Monitor write-ahead log growth and disk usage
Complete Prometheus/Grafana guide: Monitoring Integration
MIT License - Copyright (c) 2024-2025, pgElephant, Inc.
See LICENSE for details.
- Complete Documentation
- Interactive Demo - Try pg_stat_insights online
- Blog Article - Comprehensive guide and best practices
- GitHub Repository
- Report Issues
- Discussions
Built by pgElephant, Inc.
Making PostgreSQL monitoring better, one metric at a time