Skip to main content

Connection Pooling

Connection pooling is a technique where a client application maintains a set of pre-established Redis connections and reuses them instead of creating a new connection for every request.

Without pooling:

  • Application opens a new TCP connection → sends a request → closes it
  • High overhead
  • Slow
  • Causes connection churn on Redis server
  • Leads to TIME_WAIT and socket exhaustion

With pooling:

  • A pool of persistent Redis connections is created once
  • Each request picks an idle connection
  • Returns it to the pool after use

This significantly improves:

  • Latency
  • Throughput
  • Server stability
  • Concurrent request handling

Why Connection Pooling Improves Performance​

Opening/closing a TCP connection is expensive:

TCP overhead includes:

  • Handshake (SYN → SYN-ACK → ACK)
  • TLS handshake (if SSL enabled)
  • Kernel allocation for sockets
  • Redis authentication (AUTH)
  • Redis handshake (HELLO/RESP negotiation)

All of these add milliseconds per operation.

If an app performs 1,000 operations/sec, but each requires creating a connection, the overhead becomes huge.

Connection pooling eliminates all this by keeping connections alive.

Redis Server Connection Limits​

Redis is fast with commands, but slow to handle thousands of new connections per second.

Default connection limit:

maxclients 10000

If your application does not pool connections, it may exhaust maxclients under load.

Pooling minimizes client count.

How Connection Pooling Works (Architecture)​

Application
 ├── Worker Thread A ---┐
 ├── Worker Thread B ---|--> Connection Pool (10–100 connections)
 ├── Worker Thread C ---┘
 └── Worker Thread D → Redis Server

Workers borrow connections from the pool → execute → return connections.