Skip to main content

PostgreSQL

Open-source object-relational database management system. ACID-compliant with extensibility features.

On this page

What is it?

Open-source object-relational database management system. ACID-compliant with extensibility features. Supports JSON, full-text search, and rich extension ecosystem. Current versions: 14-18 (18 in preview).

Neon: Serverless PostgreSQL platform that separates compute and storage. Compute runs on Kubernetes, storage uses custom Pageserver backed by S3. Enables independent scaling and scale-to-zero.

What is it for?

  • Web applications requiring relational data with ACID guarantees
  • Development/staging with database branching (instant schema + data duplication)
  • Serverless applications with variable workloads
  • Applications needing JSONB, full-text search, extensions

Why was it chosen?

Aspect PostgreSQL MySQL SQLite
JSON Native JSONB JSON type JSON functions
Full-text search Built-in Plugin Extension
Concurrent writes MVCC Locks Single-writer
Extensions Rich ecosystem Limited Limited

Neon advantages:

  • Scale-to-zero: idle databases cost nothing
  • Instant provisioning (seconds, not minutes)
  • Database branching for testing/staging
  • Pay only for active compute time

See vendors.md for pricing, free tier limits, and provider alternatives.

Connection

Neon connections have two independent axes: endpoint type and protocol.

Endpoint type:

Endpoint Hostname Notes
Direct ep-xxx.region.aws.neon.tech Raw Postgres, required for migrations
Pooled ep-xxx-pooler.region.aws.neon.tech PgBouncer in transaction mode

Pooled (PgBouncer) disables: SET, PREPARE, LISTEN/NOTIFY, session advisory locks. Use direct for all schema operations.

Protocol:

Protocol API Use case
HTTP fetch neon() from @neondatabase/serverless Serverless — stateless, no persistent connection
WebSocket Pool/Client from @neondatabase/serverless Session persistence, interactive transactions
TCP pg, postgres.js Traditional long-lived servers

What this project uses:

  • App runtime: neon() HTTP driver + drizzle-orm/neon-http — one fetch per query, no connection pool needed, ideal for Vercel serverless
  • drizzle-kit CLI: reads DATABASE_URL, requires a direct (non-pooled) endpoint for migrations

A single direct DATABASE_URL covers both. The HTTP driver is already stateless — pooling adds nothing. drizzle-kit requires direct to run schema operations.

Known limitations

Cold starts:

  • Activating from idle: 500ms to few seconds
  • PgBouncer pooler mitigates (sub-100ms for subsequent queries)
  • Applications may timeout during reactivation

Connection limits:

  • Max connections: 112 (0.25 CU) to 4,000 (9+ CU)
  • Transaction-mode pooling disables: prepared statements (SQL-level), LISTEN/NOTIFY, SET statements, session advisory locks

Free tier constraints:

  • Compute suspends when monthly CU-hours exhausted
  • Limited storage per project
  • Short restore history and metrics retention
  • Community support only (no SLA)

See vendors.md for exact limits and paid tier features.

Related

← Back to Stack