connection-pooling

# Connection Pooling Patterns (2026)

Database and HTTP connection pooling for high-performance async Python applications.

## Overview

- Configuring asyncpg/SQLAlchemy connection pools
- Setting up aiohttp ClientSession for HTTP requests
- Diagnosing connection exhaustion or leaks
- Optimizing pool sizes for workload
- Implementing health checks and connection validation

## Quick Reference

### SQLAlchemy Async Pool Configuration

```python
from sqlalchemy.ext.asyncio import create_async_engine

engine = create_async_engine(
    "postgresql+asyncpg://user:pass@localhost/db",

    # Pool sizing
    pool_size=20,           # Steady-state connections
    max_overflow=10,        # Burst capacity (total max = 30)

    # Connection health
    pool_pre_ping=True,     # Validate before use (adds ~1ms latency)
    pool_recycle=3600,      # Recreate connections after 1 hour

    # Timeouts
    pool_timeout=30,        # Wait for connection from pool
    connect_args={
        "command_timeout": 60,      # Query timeout
        "server_settings": {
            "statement_timeout": "60000",  # 60s query timeout
        },
    },
)
```

### Direct asyncpg Pool

```python
import asyncpg

pool = await asyncpg.create_pool(
    "postgresql://user:pass@localhost/db",

    # Pool sizing
    min_size=10,            # Minimum connections kept open
    max_size=20,            # Maximum connections

    # Connection lifecycle
    max_inactive_connection_lifetime=300,  # Close idle after 5 min

    # Timeouts
    command_timeout=60,     # Query timeout
    timeout=30,             # Connection timeout

    # Setup for each connection
    setup=setup_connection,
)

async def setup_connection(conn):
    """Run on each new connection."""
    await conn.execute("SET timezone TO 'UTC'")
    await conn.execute("SET statement_timeout TO '60s'")
```

### aiohttp Session Pool

```python
import aiohttp
from aiohttp import TCPConnector

connector = TCPConnector(
    # Connection limits
    limit=100