KeyDB - A Database Built for Scale
KeyDB is a fully open source database, backed by Snap, and a faster drop in alternative to Redis
See download options available for KeyDB including Docker, DEB/RPM packages, different build options, and community provided distribution methods
Go directly to github to see source code, create issues/PRs, or if you like the project give it a star or follow us!
Check out our blog where we take an in depth look at different topics, features, and comparisons of KeyDB
KeyDB Speeds Up The User Experience For Any Project
Whether you're starting small or serving millions of users,
KeyDB enables you to provide a fast and reliable experience to your users
KeyDB is meant to handle heavy workloads with a single node benchmarking at over 1 million ops/sec. KeyDB is a multithreaded database and will outperform Redis on a per-node basis.
A Variety of Data Structures
A variety of data structures are supported such as strings, hashes, lists, sets, sorted sets, bitmaps, hyperloglogs, geospatial indexes, and streams
Multiple Persistence Options
Periodically dump the dataset to disk or by appending each command to a disk-based log. Durability preferences for RDB and AOF persistence are configurable
Scalable to Any Workload
While a single KeyDB node scales vertically, you can scale horizonally through active-replication, or cluster-mode (sharded dataset) to meet much larger workloads.
Open Source Features That Set Us Apart
While KeyDB maintains parity with the vanilla Redis feature set,
it offers some major open source advancements
MVCC Non-Blocking Architecture
With an MVCC implementation at the underlying architecture, KeyDB can query individual snapshots of the database, avoiding otherwise blocking calls such as SCAN and KEYS. Such queries can now be called concurrently at scale without reducing overall performance of existing workloads
Cross Region Multi-Master Support
Run multiple master nodes replicated asynchronously to eachother with a last write wins methodology. This enables support of a single replicated dataset where all nodes are masters with no need for sentinel monitoring nodes for failover.
KeyDB offers Subkey EXPIREs which enables expiration of members within a set. EXPIREs now also have near real time active deletion that removes major lags associated with old models of removing expired keys.
KeyDB offers TLS support that can operate at 7x the throughput of Redis + TLS
While TLS encryption adds additional CPU overhead, KeyDB’s multithreaded architecture enables more worker threads to prevent any decline in performance.