Stage 1: Problem
The Problem
You are building a backend service that connects to a PostgreSQL database. Multiple modules — UserService, OrderService, ProductService — all need database access.
typescript
| 1 | class UserService { |
| 2 | async getUser(id: string) { |
| 3 | const pool = new Pool({ host: 'localhost', port: 5432, database: 'app', max: 20 }); |
| 4 | const result = await pool.query('SELECT * FROM users WHERE id = $1', [id]); |
| 5 | pool.end(); |
| 6 | return result.rows[0]; |
| 7 | } |
| 8 | } |
| 9 | |
| 10 | class OrderService { |
| 11 | async getOrders(userId: string) { |
| 12 | const pool = new Pool({ host: 'localhost', port: 5432, database: 'app', max: 20 }); |
| 13 | const result = await pool.query('SELECT * FROM orders WHERE user_id = $1', [userId]); |
| 14 | pool.end(); |
| 15 | return result.rows; |
| 16 | } |
| 17 | } |
Each service creates its own connection pool, connects, queries, and disconnects. This works for a single request. But now your app handles 500 requests per second.
What happens? Each request creates a new pool → new TCP connections → handshakes → authentication → query → teardown. You hit PostgreSQL's max_connections limit (default 100). The database starts refusing connections. Your app crashes under load.
Think about it: How would you ensure all services share a single connection pool, initialized once, without passing it everywhere manually?