pressure

A simple, java-based, http server stress test tool, git address

Cause

jmeter dynamic scripting is more complicated
Hope that the stress test program can be executed on the server side, and initiate a request from the intranet environment
There is enough room for expansion to facilitate pressure testing of complex processes

rely

JDK 8
Maven

Quick start

Construct

Execute git command

git clone https://github.com/NotInWine/pressure.git

Import your idea
Build the project using pom.xml
```
步骤略过
```

Examples

package com.pressure;

/**
 * 测试
 *
 * @author YL
 */
public class Main {

    public static void main(String[] args) {

        /**
         * 获取请求处理器
         * 总请求次数等于循环次数 * RequestInfo 脚本长度
         * @param poolSize 线程池并发线程数量
         * @param requestSize 循环请求次数
         */
        RequestInterface requestInterface = RequestImpl.build(
                500,
                5000)
                .setContentTimeOut(3000) // 链接超时时间 毫秒
                .setSocketTimeOut(5000)  // 响应超时时间 毫秒
                .setMonitors(
                        /**
                         * 设置监听（收集统计压测记录）
                         * 目前自带两个款监听器
                         * @see com.pressure.core.httputil.impl.LogResultInfoMonitor 日志记录
                         * @see com.pressure.core.httputil.impl.SummaryMonitor 汇总打印
                         *
                         * 想要实现自定义的监听器可实现:
                         * @see com.pressure.core.httputil.ResultInfoMonitor
                         * 在此处配置给处理器
                         */
                        new LogResultInfoMonitor(new File("D:\\big_folder\\log.txt")),
                        new SummaryMonitor()
                );

        // 配置请求脚本
        requestInterface.send(
                new RequestInfo(
                        "列表",
                        "https://www.xxxxxx.com/api/video/pay/start?start=1&count=15&keyWord=",
                        RequestInfo.RequestMethod.GET
                ),
                new RequestInfo(
                        "详情(包含动态参数)",
                        "https://www.xxxxxx.com/api/video/playDetail?videoId=${data.data.$2.id}&token=",
                        RequestInfo.RequestMethod.GET
                ),
                new RequestInfo(
                        "子列表(包含动态参数)",
                        "https://www.xxxxxx.com/api/video/getVideList?recordId=${data.videoId}&type=video",
                        RequestInfo.RequestMethod.GET
                ),
                new RequestInfo(
                        "子列表详情(包含动态参数)",
                        "https://www.xxxxxx.com/api/video/playDetail?videoId=${data.$1.id}&token=",
                        RequestInfo.RequestMethod.GET
                )
        );
    }
}

Sample output

LogResultInfoMonitor log

batchId=4752, name=子列表(包含动态参数), state=SUCCESS, httpLog=HttpLog{url='https://xkx.aiyingli.com/api/video/getVideList?recordId=245&type=video', httpState=200, params=null
, requestHeaders=null
, responseHeaders=[Server: nginx, Date: Sun, 19 Apr 2020 04:59:12 GMT, Content-Type: application/json; charset=utf-8, Transfer-Encoding: chunked, Connection: keep-alive, X-Powered-By: PHP/5.5.38, Access-Control-Allow-Origin: *]
, responseBody='{"code":1,"msg":"success","time":"1587272352","data":[{]}'}
, time=4260

SummaryMonitor

子列表:{, 最慢响应=9121, 最快响应=57, 平均响应=2533, 总请求次数=4072, 请求异常次数=301, 异常率=0.0739}
详情:{, 最慢响应=5819, 最快响应=56, 平均响应=2501, 总请求次数=4438, 请求异常次数=366, 异常率=0.0825}
列表:{, 最慢响应=10718, 最快响应=48, 平均响应=2862, 总请求次数=5000, 请求异常次数=562, 异常率=0.1124}
子列表详情:{, 最慢响应=5086, 最快响应=54, 平均响应=2489, 总请求次数=3771, 请求异常次数=272, 异常率=0.0721}
汇总:{吞吐量(qps)=189.90, 最慢响应=10718, 最快响应=48, 平均响应=2610, 总请求次数=17281, 请求异常次数=1501, 异常率=0.0869}

Package execution

Need to be packaged to the server to execute the editable example , and then use mvn install, marked as an executable jar, and execute it on (server | PC) java -jar

Key concept

The controller RequestInterface
is responsible for the core logic of the pressure test request, responsible for http connection configuration, request sending, and trigger statistics (calling the listener)
The request command package RequestInfo
encapsulates the request instruction, which is passed as a parameter to the controller for execution.
The listener ResultInfoMonitor
handles the pressure test results and currently supports two listeners, receiving response logs
- Log
- Summary monitoring
  As the most scalable part, you can use the listener to customize various forms of statistical output