cloudflare-hyperdrive

Connection pooling with Hyperdrive lets you connect Workers to PostgreSQL and MySQL databases using edge-based global pooling. Configure a Hyperdrive instance pointing to your database, then query it from your Worker using node-postgres, postgres.js, mysql2, Drizzle ORM, or Prisma.

Connection pooling reuses database connections across requests instead of creating new ones, eliminating 7 round trips (TCP, TLS, authentication). Hyperdrive's edge pooling significantly reduces latency and connection overhead when Workers query databases.

Yes, Hyperdrive works with existing PostgreSQL and MySQL databases including AWS RDS instances. Configure Hyperdrive with your database credentials and connection details, then connect your Worker to query data immediately without modifying your existing database setup.

Hyperdrive supports prepared statements for query caching and integrates with major ORM drivers including Drizzle ORM and Prisma, plus database libraries like node-postgres and mysql2. This enables efficient, cacheable queries from Workers.

Hyperdrive enforces a 5-connection pool cap per Worker. Use prepared statements and ctx.waitUntil for non-blocking cleanup. TLS/SSL enforcement is required; avoid eval() statements since they're disallowed in Worker environments.

Pool acquisition failures and TLS errors typically arise from misconfigured credentials, connection limits, or TLS settings. Verify your Hyperdrive configuration matches your database requirements, ensure TLS/SSL is properly enabled, and check that your Worker respects the 5-connection pool cap.