Skip to main content
Connect PostgreSQL databases to Tembo for performance monitoring and optimization.

​
Features

  • Monitor slow queries - Identifies queries that exceed performance thresholds and suggests optimizations
  • Detect missing indexes - Analyzes query patterns to find opportunities for new indexes that would improve performance
  • Detect unused indexes - Identifies indexes that are no longer being used, helping reduce storage overhead and write performance impact
  • Automatic optimization PRs - Generates pull requests with database migration scripts to implement recommended improvements
  • Task-based data access - Postgres MCP enables Tembo to query your database when needed for specific tasks (always read-only)

​
Installation

1

Create User (Optional)

Optional: Create a dedicated database user for Tembo to ensure read-only access. While not strictly necessary, this is recommended for additional security assurance. The user needs access to PostgreSQL system catalogs and statistics views.Example SQL:
CREATE USER tembo_monitor WITH PASSWORD 'your-secure-password';
GRANT CONNECT ON DATABASE your_database TO tembo_monitor;
GRANT USAGE ON SCHEMA public TO tembo_monitor;
GRANT SELECT ON ALL TABLES IN SCHEMA public TO tembo_monitor;
2

Connect

Navigate to the Integrations page in Tembo and click the Connect button next to Postgres. Enter your connection string in the format: postgresql://username:password@host:5432/database
3

Test Connection

Test the connection to verify Tembo can access your database. Once verified, monitoring begins immediately.
4

Map Repository

Map your database connection to the GitHub, GitLab, or Bitbucket repository where your database migrations are stored. This allows Tembo to create PRs with optimization suggestions.

​
Usage

​
Automated Monitoring

Once connected, Tembo continuously monitors your database performance:
  • Query Performance Analysis - Tracks slow queries using PostgreSQL’s pg_stat_statements extension
  • Index Analysis - Examines query execution plans to identify missing or unused indexes
  • Performance Trends - Monitors query performance over time to detect degradation

​
Pull Requests for Optimizations

When Tembo identifies optimization opportunities, it creates a PR with migration scripts (e.g., CREATE INDEX), explains the performance issue, and shows expected improvement metrics.

​
Postgres MCP Integration

Tembo’s Postgres MCP may access your database during task execution when querying it would help complete a task. It runs in restricted mode (read-only), uses your configured connection, and notifies you when accessed.

​
Best Practices

  • Use Read-Only Access - Tembo only needs read access to monitor performance; it never modifies your database directly
  • Monitor Production Databases - Connect production databases for the most accurate performance insights
  • Review Migration Scripts - Always review Tembo’s suggested migrations before merging, especially for large tables
  • Test in Staging - Apply migrations to a staging environment first to verify performance improvements
  • Enable pg_stat_statements - Ensure the pg_stat_statements extension is enabled for comprehensive query analysis. Enable it by running CREATE EXTENSION IF NOT EXISTS pg_stat_statements; in your database. If the extension isn’t available, add pg_stat_statements to shared_preload_libraries in your postgresql.conf and restart PostgreSQL.