Dragonfly
A high-performance, thread-per-core in-memory data store designed for extreme throughput and scalability, fully compatible with Redis APIs.
Dragonfly is an open-source alternative to:
Dragonfly
Dragonfly is a modern, high-performance in-memory data store engineered for extreme throughput at scale. Built with a thread-per-core architecture, it eliminates the scaling bottlenecks of traditional real-time data infrastructure like Redis, delivering lower latency, higher throughput, and better memory efficiency on a smaller hardware footprint.
Key Features
- Thread-per-core architecture: Leverages every CPU core for maximum parallelism and performance.
- Redis-compatible: Drop-in replacement for Redis with full API compatibility.
- Extreme throughput: Achieves up to 25x higher QPS than Redis in benchmarks.
- Memory efficiency: Uses up to 30% less memory compared to Redis.
- Scalability: Single instance handles workloads typically requiring a Redis cluster.
- Cloud-native: Optimized for modern cloud hardware and infrastructure.
- Multi-threaded: Eliminates single-threaded bottlenecks found in Redis.
- Use cases: High-performance caching, real-time feature stores, message queues, and leaderboards.
Tech / Self-Hosting
Dragonfly is open-source and can be self-hosted on any Linux system. It is also available as a managed cloud service (Dragonfly Cloud) with a free tier offering $100 in credits. The project is actively maintained with a growing community and extensive documentation.
GitHub Repository: https://github.com/dragonflydb/dragonfly
Managed Service: https://dragonflydb.cloud/
Documentation: https://www.dragonflydb.io/docs