Stress testing tool WRK

 

 

Project startup settings: java -Xmx4g -Xms4g -jar *.jar

 

when testing

   Execute: ./wrk -t 5 -c100 -d 6 --latency -T 40 http://192.168.1.206:8004/qixiangControler/queryData?currentPage=1

   Use 5 threads for each case, and 100, 200, 500, 1000 for concurrency testing

 

 

 

The stress test wrk tool uses:

Source address: https://github.com/wg/wrk

Install:

   

git clone https://github.com/wg/wrk.git  
cd wrk  
make  

 

After the installation is complete, perform a stress test:

     

./wrk -t 5 -c19   -d 6 --latency -T 40   http://192.168.1.206:8004/qixiangControler/queryData?currentPage=1

 

Parameter Description:

     

    -c, --connections <N> Connections to keep open number of connections, concurrent
    -d, --duration <T> Duration of test stress test duration, that is, in how many seconds the stress test will be performed, and it will end automatically after the time expires      
    -t, --threads <N> Number of threads to use Number of threads to use, set to 2-3 times the combined number of cpu
                                                      
    -s, --script      <S>  Load Lua script file       
    -H, --header      <H>  Add header to request      
        --latency Print latency statistics Print latency distribution, that is, the corresponding time distribution
        --timeout <T> Socket/request timeout The default timeout for each thread is 1 second
    -v, --version          Print version details   

 

 

Practical example: The following is a stress test of an interface with 5 threads and 19 concurrent

 ./wrk -t 5 -c19   -d 6 --latency -T 40   http://192.168.1.206:8004/qixiangControler/queryData?currentPage=1

    

Running 6s test @ http://192.168.1.206:8004/qixiangControler/queryData?currentPage=1
  5 threads and 19 connections
  Thread StatsAvg    Stdev Max  +/- Stdev
    Latency response time 5.33s Average response time 343.23ms 5.75s Maximum response time 60.00%
    Req/Sec Throughput 2.20 Average Throughput 3.43 10.00 Maximum Throughput 90.00%
  Latency Distribution
     50% 5.34s
     75% 5.73s
     90% 5.75s
     99% 5.75s

 

  15 requests in 6.01s, 21.53MB read, (processing 15 requests in 6 seconds, and processing 2.5 requests per second, the total throughput is 2.5)

Requests/sec: 2.50 total throughput

Transfer/sec:      3.58MB

Increasing the -c parameter is the number of concurrency, and the throughput is reduced, so the maximum concurrency is 19

 

 

2. Apache Jmeter graphical stress testing tool