Dragonfly
Drop-in Redis replacement with up to 25x throughput on a modern engine.
About
Dragonfly is a Redis-compatible in-memory datastore rebuilt on a shared-nothing, thread-per-core architecture. It speaks the Redis RESP2/RESP3 protocols natively, so existing clients like ioredis, redis-py, and go-redis work with no code changes, while a single instance can serve millions of operations per second by using every CPU core instead of Redis's single-threaded model.
Founded in 2022 by ex-Google engineers Oded Poncz and Roman Gershman, DragonflyDB raised $21M across a Redpoint-led seed and a Quiet Capital-led Series A. Independent benchmarks typically show 8-15x Redis throughput on mixed workloads, and users like Instacart have collapsed sharded Redis clusters into far smaller vertically scaled deployments. The license is BSL 1.1 (source-available, not OSI-approved), and Dragonfly Cloud offers a managed alternative to Redis Cloud.
What it can do
- Full Redis protocol compatibility with zero client code changes
- Multi-threaded engine scaling linearly with CPU cores
- Vertical scaling that collapses sharded clusters and cuts cost
- SSD tiering and a managed cloud offering
Setup
docker run -p 6379:6379 dragonflydb/dragonfly, then point your existing Redis client at it. Or use Dragonfly Cloud managed.
Prerequisites
An existing Redis workload or a new caching need; no client changes required.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 98; Trending score 80/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.