facebook/rocksdb

Health v2 · maintenance only

Adoption not part of the score

32197 stars · 6959 forks observed · 2026-10-14

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

RocksDB is an embeddable persistent key-value store library written in C++, based on LevelDB and maintained by Meta's Database Engineering team. It uses an LSM-tree design optimized for fast storage like flash drives and RAM, supporting terabytes of data with tunable write, read, and space amplification tradeoffs.

Use cases

• embed a fast key-value store in my application
• store large amounts of data on flash drives with low latency
• build a database storage engine
• use an LSM-based persistent store for write-heavy workloads
• cache application data persistently
• store multiple terabytes in a single embedded database

When to choose

• you need an embedded, persistent key-value store with high write throughput
• your workload runs on flash or RAM storage and needs low latency
• you want tunable amplification tradeoffs and multi-threaded compaction
• you're building a server or engine that needs a proven storage core

When to avoid

• you need a standalone client-server database rather than an embedded library
• you need SQL or rich querying - RocksDB only stores ordered byte keys and values
• you want a simple managed database without C++ integration work
• GPLv2/Apache dual licensing is incompatible with your project

Facets

library · maturity stable

database caching databases developer-tools backend cpp windows cross-platform key-value-store lsm-tree storage-engine embedded-database leveldb flash-storage linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="facebook/rocksdb")