Fix busy event emission on fixed pool: (#332)
[poolifier.git] / benchmarks / README.md
1 # Poolifier Benchmarks
2
3 Welcome to poolifier benchmarks and thanks to look into this project.
4
5 ## Folder Structure
6
7 The internal folder contains poolifier internal benchmarks.
8 The versus-external-pools folder contains benchmarks versus other Node.js pools.
9
10 ## Poolifier vs other pools benchmark
11
12 To compare poolifier pools performance vs other pools performance we chose to use [hyperfine](https://github.com/sharkdp/hyperfine).
13 We chose to use this tool because it allows to run isolated Node.js processes so that each pool does not impact each other.
14 External pools with which we compared the poolifier results:
15
16 - [piscina](https://github.com/piscinajs/piscina)
17 - [SUCHMOKUO/node-worker-threads-pool](https://github.com/SUCHMOKUO/node-worker-threads-pool)
18 - [threads.js](https://github.com/andywer/threads.js/)
19 - [workerpool](https://github.com/josdejong/workerpool)
20 - [threadwork](https://github.com/kevlened/threadwork)
21 - [microjob](https://github.com/wilk/microjob)
22 - [worker-threads-pool](https://github.com/watson/worker-threads-pool)
23
24 Those are our results:
25
26 - CPU Intensive task with 100k operations submitted to each pool [BENCH-100000.md](./versus-external-pools/BENCH-100000.md).
27 This benchmark ran on a MacBook Pro 2015, 2,2 GHz Intel Core i7 quad-core, 16 GB 1600 MHz DDR3.
28
29 > :warning: **We would need funds to run our benchmarks more often and on Cloud VMs, please consider to sponsor this project**
30
31 ## How to run benchmarks
32
33 ### Internal
34
35 To run the internal benchmark you just need to navigate to the root of poolifier project and run `npm run benchmark`
36
37 ## Versus other pools
38
39 To run the benchmark versus other pools you will need to:
40
41 - [Install hyperfine](https://github.com/sharkdp/hyperfine#installation)
42 - Run the `./bench.sh` into the `versus-external-pools` folder
43
44 > :warning: **Please be sure to use a quite PC when you run the benchmarks**