http-stats
v2.0.0
Published
Module for benchmarking HTTP servers with support for metrics such as response time, requests per second and CPU and memory usage. Generates friendly HTML reports and offers both a command line interface and a JavaScript API.
Downloads
14
Readme
http-stats
Simple tool for load testing an HTTP server that allows for increasing load to be applied over time. This module provides support for collecting various statistics, including requests per second, response time, CPU and memory usage. Resulting statistics can be saved to disk as a JSON file or written to an HTML report with graphs.
Installation
To install the command line tool:
npm install http-stats --globalAlternatively, the JavaScript module can be installed into a project:
npm install http-stats --saveUsage
Command Line Interface
http-stats --url http://localhost:8080 -n 100 --begin-concurrency 1 --end-concurrency 50 --pids 1234 1235 --report-dir reportAlternatively, all of the configuration can be put into a configuration file:
http-stats --config my-config.jsonExample contents of my-config.json:
{
"url": "http://localhost:8080",
"beginConcurrency": 1,
"endConcurrency": 50,
"stepRequests": 100,
"pids": [1234, 1235],
"report": {
"outputDir": "path:./report"
}
}JavaScript API
Alternatively, the JavaScript API can be used to collect statistics as shown in the following sample code:
var httpStats = require('http-stats');
httpStats.measure({
url: 'http://localhost:8080',
beginConcurrency: 1,
endConcurrency: 50,
stepRequests: 100,
pids: [1234, 1235],
report: {
outputDir: require('path').join(__dirname, 'report')
}
}, function(err, results) {
// Do something with the resulting data
});Sample Statistics
{
"cpuAvailable": true,
"memoryAvailable": true,
"steps": [
{
"concurrency": 1,
"stats": {
"responseTime": {
"min": 11.562785029411316,
"max": 19.471349954605103,
"sum": 1358.6313557624817,
"variance": 2.64784231895231,
"mean": 13.586313557624816,
"stddev": 1.627219198188219,
"count": 100,
"median": 13.028658986091614,
"p75": 14.24703973531723,
"p95": 16.65775092244148,
"p99": 19.463981754779812,
"p999": 19.471349954605103
},
"requestsPerSecond": {
"mean": 73.32546082534017,
"count": 100,
"currentRate": 73.32433734563027,
"1MinuteRate": 0,
"5MinuteRate": 0,
"15MinuteRate": 0
},
"usage": [
{
"pid": "33062",
"cpu": {
"min": 0,
"max": 0.5,
"sum": 11.199999999999989,
"variance": 0.022248995983935746,
"mean": 0.1333333333333332,
"stddev": 0.14916097339430226,
"count": 84,
"median": 0.1,
"p75": 0.2,
"p95": 0.5,
"p99": 0.5,
"p999": 0.5
},
"memory": {
"min": 68.20703125,
"max": 68.20703125,
"sum": 5797.59765625,
"variance": 0,
"mean": 68.20703125,
"stddev": 0,
"count": 85,
"median": 68.20703125,
"p75": 68.20703125,
"p95": 68.20703125,
"p99": 68.20703125,
"p999": 68.20703125
}
}
]
}
},
{
"concurrency": 2,
"stats": {
"responseTime": {
"min": 11.659276962280273,
"max": 21.81683897972107,
"sum": 1383.8951340913773,
"variance": 3.2013560905275846,
"mean": 13.97873872819573,
"stddev": 1.7892333806766474,
"count": 99,
"median": 13.467664003372192,
"p75": 14.689211010932922,
"p95": 17.55275595188141,
"p99": 21.81683897972107,
"p999": 21.81683897972107
},
"requestsPerSecond": {
"mean": 141.74595662204692,
"count": 99,
"currentRate": 141.73979759785095,
"1MinuteRate": 0,
"5MinuteRate": 0,
"15MinuteRate": 0
},
"usage": [
{
"pid": "33062",
"cpu": {
"min": 0,
"max": 0.3,
"sum": 5.9,
"variance": 0.012024390243902443,
"mean": 0.14390243902439026,
"stddev": 0.1096557807135695,
"count": 41,
"median": 0.2,
"p75": 0.2,
"p95": 0.3,
"p99": 0.3,
"p999": 0.3
},
"memory": {
"min": 68.20703125,
"max": 68.20703125,
"sum": 2864.6953125,
"variance": 0,
"mean": 68.20703125,
"stddev": 0,
"count": 42,
"median": 68.20703125,
"p75": 68.20703125,
"p95": 68.20703125,
"p99": 68.20703125,
"p999": 68.20703125
}
}
]
}
},
...,
{
"concurrency": 10,
"stats": {
"responseTime": {
"min": 11.175518989562988,
"max": 29.486907958984375,
"sum": 1753.7954921722412,
"variance": 21.318200618972394,
"mean": 19.272477935958694,
"stddev": 4.617163698524495,
"count": 91,
"median": 19.485509037971497,
"p75": 21.68475103378296,
"p95": 28.69777555465698,
"p99": 29.486907958984375,
"p999": 29.486907958984375
},
"requestsPerSecond": {
"mean": 504.8386172505691,
"count": 91,
"currentRate": 504.8419813195869,
"1MinuteRate": 0,
"5MinuteRate": 0,
"15MinuteRate": 0
},
"usage": [
{
"pid": "33062",
"cpu": {
"min": 0,
"max": 0.4,
"sum": 2.4,
"variance": 0.022857142857142864,
"mean": 0.34285714285714286,
"stddev": 0.1511857892036909,
"count": 7,
"median": 0.4,
"p75": 0.4,
"p95": 0.4,
"p99": 0.4,
"p999": 0.4
},
"memory": {
"min": 68.2109375,
"max": 68.2109375,
"sum": 545.6875,
"variance": 0,
"mean": 68.2109375,
"stddev": 0,
"count": 8,
"median": 68.2109375,
"p75": 68.2109375,
"p95": 68.2109375,
"p99": 68.2109375,
"p999": 68.2109375
}
}
]
}
}
],
"url": "http://localhost:9900",
"pids": [
33062
],
"options": {
"beginConcurrency": 1,
"endConcurrency": 10,
"stepRequests": 100,
"concurrencyIncrement": 1,
"delay": 2000
}
}TODO
- Add documentation for all supported configuration options
- Integrate support for pm2 to query for process IDs using the pm2-interface module.
- Provide ability to compare differences between statistics gathered at different times (to detect performance decreases/increases)
- Improve tests
- Provide link to sample HTML report from this documentation
Maintainers
- Patrick Steele-Idem (Github: @patrick-steele-idem) (Twitter: @psteeleidem)
License
Credits
- The reports generated by this tool use Highcharts to render the graphs.
- The measured module is used to collect statistics.