A modern lock-free atomic embedded database designed to beat LSM trees for reads and traditional B+ trees for writes.
It uses a modular design which can also be used to implement your own high
performance persistent systems, using the included
Eventually, a versioned DB will be built on top of the
Tree which provides
multi-key transactions and snapshots.
Tree has a C API, so you can use this from any mainstream language.
extern crate sled; let tree = sled::Config::default() .path(Some(path)) .tree(); // set and get tree.set(k, v1); assert_eq!(tree.get(&k), Some(v1)); // compare and swap tree.cas(k, Some(v1), Some(v2)); // scans let mut iter = tree.scan(b"a non-present key < k!"); assert_eq!(iter.next(), Some((k, v2))); assert_eq!(iter.next(), None); // deletion tree.del(&k);
- Quite young, there are lots of fuzz tests but don't bet a billion dollar business on it yet!
- The C API is likely to change rapidly
- Log cleaning is currently only implemented for linux via
- Has not yet received much attention for performance tuning, it has an extremely high theoretical performance but there is a bit of tuning to get there. Currently only around 200k operations per second with certain contrived workloads. This will be improving soon!
- Want to help advance the state of the art in open source embedded databases? Check out CONTRIBUTING.md!
- lock-free b-link tree
- lock-free log with reservable slots
- lock-free pagecache with cache-friendly partial updates
- zstd compression
- configurable cache size
- C API
- log cleaning
- merge operator support
- higher-level interface with multi-key transaction and snapshot support
- formal verification of lock-free algorithms via symbolic execution
- beat LSM's on read performance and traditional B+ trees on write performance.
- don't use so much electricity. our data structures should play to modern hardware's strengths.
- don't surprise users with performance traps.
- bring reliability techniques from academia into real-world practice.
Lock-free trees on a lock-free pagecache on a lock-free log. The pagecache scatters partial page fragments across the log, rather than rewriting entire pages at a time as B+ trees for spinning disks historically have. On page reads, we concurrently scatter-gather reads across the log to materialize the page from its fragments.
The system is largely inspired by the Deuteronomy architecture, and aims to implement the best features from RocksDB as well.