Dragonfly

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

31779 stars · 1279 forks observed · 2026-10-12

What it is AI-extracted, prompt v1, taxonomy v1, 2026-10-05, confidence 0.97

Dragonfly is a modern, multi‑threaded in‑memory data store that implements the Redis and Memcached APIs. It provides a drop‑in replacement for those systems with dramatically higher throughput and lower latency.

Use cases

• migrate a Redis cache to achieve higher QPS without code changes
• run a high‑performance feature store for machine‑learning models
• serve session data for web applications with sub‑millisecond latency
• implement a job‑queue backend that needs millions of operations per second
• replace Memcached in a microservices architecture to reduce hardware costs

When to choose

• you need Redis/Memcached compatibility but want significantly better performance
• your workload is CPU‑bound and can benefit from a thread‑per‑core architecture
• you want a single‑binary deployment that runs on Docker, Kubernetes, or bare metal
• you aim to reduce memory footprint and operational costs while scaling vertically

When to avoid

• you require multi‑node clustering or automatic sharding across many machines (Dragonfly is primarily single‑node vertical scaling)
• your environment only supports languages without a Redis client library
• you need built‑in persistence guarantees comparable to traditional relational databases

Facets

application · maturity active

database caching databases cloud-computing windows cloud self-hosted redis-compatible memcached-compatible multi-threaded high-throughput in-memory kv-store vertical-scaling caching devops linux macos docker kubernetes

Member repositories

For agents

markdown · JSON · MCP: product_card(name="dragonflydb/dragonfly")