jmeter静默压测+可视化

静默压测自动化脚本auto_stress_test.sh

#!/usr/bin/env bash
export jmx_template="test2"
export suffix=".jmx"
export jmx_template_filename="${jmx_template}${suffix}"
export os_type=`uname`

# 需要在系统变量中定义jmeter根目录的位置，如下
#export jmeter_path=/usr/local/jmeter/apache-jmeter-5.2

# 清空nohup.out
cat /dev/null > nohup.out

# 强制杀掉JMeter进程
killJMeter()
{
    pid=`ps -ef|grep jmeter|grep java|grep ${jmx_filename}|awk '{print $2}'`
    echo "jmeter Id list :$pid"
    if [[ "$pid" = "" ]]
    then
      echo "no jmeter pid alive"
    else
      kill -9 $pid
    fi
}

thread_number_array=(100 200 300)
for num in "${thread_number_array[@]}"
do
    # 生成对应压测线程的jmx文件
    export jmx_filename="${jmx_template}_${num}${suffix}"
    export jtl_filename="test_${num}.jtl"

    rm -f ${jmx_filename} ${jtl_filename}
    cp ${jmx_template_filename} ${jmx_filename}
    echo "生成jmx压测脚本 ${jmx_filename}"

    if [[ "${os_type}" == "Darwin" ]]; then
        sed -i "" "s/thread_num/${num}/g" ${jmx_filename}
    else
        sed -i "s/thread_num/${num}/g" ${jmx_filename}
    fi

    # JMeter 静默压测
    nohup ${jmeter_path}/bin/jmeter -n -t ${jmx_filename} -l ${jtl_filename} &
    sleep 30
    killJMeter
    rm -f ${jmx_filename}
done
echo "自动化压测全部结束"

jmeter脚本test2.jmx必须和auto_stress_test.sh脚本在同一个文件夹下

docker 安装influxdb

启动influxdb
docker run -d -p 8086:8086 -p 8083:8083 --name=jmeterdb influxdb

进入influxdb容器
docker exec -it jmeterdb bash

进入influxdb输入
influx

创建数据库，名称为：jmeter
create database jmeter;

进入创建的数据库
use jmeter