Database
/
Redis Storage

Redis Storage

Redis primarily serves as a high-performance message broker for streaming data in Open LangGraph Server. While Redis can also be used for checkpoints when needed, its main role is handling real-time pub/sub messaging for graph execution streams.

Overview

Redis is used in Open LangGraph Server mainly for streaming data synchronization through pub/sub messaging. It can optionally be used for checkpoint storage, but this is typically not recommended for production use due to data persistence concerns.

Characteristics

  • 📡 Message Broker - Primary role for streaming data
  • High performance - Sub-millisecond pub/sub operations
  • 🔄 Pub/Sub messaging - Real-time data synchronization
  • 💾 Optional checkpoint storage - Available but not recommended for production

Configuration

Basic Configuration

Optional Checkpoint Storage

Redis can be used for checkpoint storage (not recommended for production):

Advanced Configuration

Checkpoint Storage (Optional)

While Redis is primarily used for message queues, you can optionally use it for checkpoint storage:

Note: Redis checkpoints are not recommended for production due to data persistence limitations. Use PostgreSQL or other persistent storage for checkpoints in production.

Open LangGraph Server Redis Keys

When using Redis, Open LangGraph Server stores data under these key patterns:

Security

Authentication

TLS/SSL

Performance Characteristics

MetricRedis Performance
Read Latency
0.1-1ms
Write Latency
0.1-1ms
Throughput
100k-1M ops/sec
Memory Usage
Configurable (256MB-1TB+)
Persistence
Optional (RDB/AOF)

Use Cases

Primary use in Open LangGraph Server:

  • Message Queues - Real-time streaming data for graph executions
  • Pub/Sub messaging - Synchronize data between graph components
  • Real-time updates - Stream execution results to clients

Optional checkpoint storage:

  • Development/Testing - Fast checkpoint access for debugging
  • High-performance caching - When persistence is not critical
  • Temporary state - Short-lived checkpoint data

Note: For production checkpoint storage, use PostgreSQL instead of Redis due to data persistence requirements.

Performance Issues

Best Practices

  • Use authentication - Always configure passwords for Redis
  • Enable TLS - Use encrypted connections in production
  • Monitor connections - Track connection pool usage
  • Set appropriate TTL - Configure proper key expiration times
Built with

Show your support! Star us on GitHub ⭐️