con-art-rust

Crates.iocon-art-rust
lib.rscon-art-rust
version0.2.0
sourcesrc
created_at2022-01-12 19:32:08.408058
updated_at2022-02-15 01:20:12.002926
descriptionA Rust implementation of ART-OLC concurrent adaptive radix tree.
homepage
repositoryhttps://github.com/XiangpengHao/con-art-rust
max_upload_size
id512978
size100,904
Xiangpeng Hao (XiangpengHao)

documentation

README

Concurrent ART (adaptive radix tree)

con-art-rust Crates.io dependency status codecov Documentation

A Rust implementation of ART-OLC concurrent adaptive radix tree. It implements the optimistic lock coupling with proper SIMD support.

It only supports (and is optimized for) 8 byte key; due to this specialization, this ART has great performance -- basic operations are ~40% faster than flurry hash table, range scan is an order of magnitude faster.

The code is extensively tested with {address|leak} sanitizer as well as libfuzzer.

Why this library?

  • Fast performance, faster than most hash tables.

  • Concurrent, super scalable, it reaches 150Mop/s on 32 cores.

  • Super low memory consumption. Hash tables often have exponential bucket size growth, which often lead to low load factors. ART is more space efficient.

Why not this library?

  • Not for arbitrary key size. This library only supports 8 byte key.
  • The value must be a valid, user-space, 64 bit pointer, aka non-null and zeros on 48-63 bits.
  • Not for sparse keys. ART is optimized for dense keys, if your keys are sparse, you should consider a hashtable.

Example:

use con_art_rust::Art;
let art = Art::new();
let guard = art.pin(); // enter an epoch

art.insert(0, 42, &guard); // insert a value
let val = art.get(&0).unwrap(); // read the value
assert_eq!(val, 42);

let mut scan_buffer = vec![0; 8];
let scan_result = art.range(&0, &10, &mut art_scan_buffer); // scan values
assert_eq!(scan_result.unwrap(), 1);
Commit count: 290

cargo fmt