@anyroad/react-component-benchmark

React Component Benchmark

This project aims to provide a method for gathering benchmarks of component tree mount, update, and unmount timings.

Please note that the values returned are estimates. Since this project does not hook into the React renderer directly, the values gathered are not 100% accurate and may vary slightly because they're taken from a wrapping component. That being said, running a large sample set should give you a confident benchmark metric.

Motivation

Historically, React has provided react-addons-perf in order to help gain insight into the performance of mounting, updating, and unmounting components. Unfortunately, as of React 16, it has been deprecated. Additionally, before deprecation, it was not usable in production React builds, making it less useful for many applications.

Usage

See the examples directory for ideas on how you might integrate this into your own project, whether in your user-interface or your automated tests.

Quick Start

import Benchmark, { BenchmarkType } from 'react-component-benchmark';

class MyComponentBenchmark extends React.Component {
  render() {
    return (
      <div>
        <button onClick={this._start}>Run</button>
        <Benchmark
          component={MyComponent}
          componentProps={componentProps}
          onComplete={this._handleComplete}
          ref={this._setBenchRef}
          samples={50}
          timeout={10000}
          type={BenchmarkType.MOUNT}
        />
      </div>
    );
  }

  _start = () => { this._benchmark.start(); };

  _handleComplete = (results) => {
    console.log(results);
  };

  _setBenchRef = (ref) => { this._benchmark = ref; };
}

Benchmark props

• component: typeof React.Component
• componentProps: object Properties to be given to component when rendering
• onComplete: (x: BenchResultsType) => void
• samples: number Samples to run (default 50)
• timeout: number Amount of time in milliseconds to stop running (default 10000)
• type: string One of 'mount', 'update', or 'unmount'. Also available from BenchmarkType.

In tests

See examples/tests for various test integrations.

Results

Note: All times and timestamps are in milliseconds. High resolution times provided when available.

• startTime: number Timestamp of when the run started
• endTime: number Timestamp of when the run completed
• runTime: number Amount of time that it took to run all samples.
• sampleCount: number The number of samples actually run. May be less than requested if the timeout was hit.
• samples: Array<{ start, end, elapsed }>
• max: number Maximum time elapsed
• min: number Minimum time elapsed
• median: number Median time elapsed
• mean: number Mean time elapsed
• stdDev: number Standard deviation of all elapsed times
• p70: number 70th percentile for time elapsed: mean + stdDev
• p95: number 95th percentile for time elapsed: mean + (stdDev * 2)
• p99: number 99th percentile for time elapsed: mean + (stdDev * 3)