Jmeter distributed deployment process

1. Why is it distributed?

  Most of the projects that require performance testing are often large-scale projects, or projects with higher requirements for interface response speed and stability, such as e-commerce, medical, live broadcast, etc., and these projects require a large number of concurrency. Stress testing can expose some problems, so at this time, the test execution machine is stuck in the first level. At this time, you need to use distributed deployment jmeter, and use multiple machines to jointly pressurize the server under test to achieve the purpose of performance testing. .

2. Distributed deployment method

1. Use tools such as xshell to connect the control machine and the execution machine

  1. Use ifconfig to obtain the IP address of the execution machine (usually a Linux system, usually multiple)

  2. Connect using tools such as xshell

    

  3. Use the xftp that comes with xshell to share the local jmeter, or you can use the command line to copy it directly, the method is not unique

    

 

   4. Modify the jmeter.properties file in the executor and control machine apache-jmeter-5.1.1\bin respectively, just change the following two places

    

    

 

 

    

 

 

     

 

   5. Start jmeter-server on the execution machine respectively

    

 

   6. The control machine starts jmeter, and you can see that it has been configured successfully. If you choose to start all remotely, all the actuators will run together.

    

 

        

    

3. Matters needing attention

  1.jmeter operation requires jdk environment, Windows installation and configuration jdk will not be repeated here, Linux needs to be copied to a certain directory, use tar -zxvf package name to decompress, and then use vi /etc/profile command to edit the configuration file

            

            

 

   2.The number of threads and cycle time set by jmeter are single. When counting, all executed scripts are counted. If two are executed, *2, you need to pay attention to it during work.

 

   3. When jmeter prompts the execution machine to reject the connection, use the firewall-cmd --state command to check whether the firewall is off. Distributed deployment needs to turn off the firewall. Use the command systemctl stop firewalld to temporarily turn off the firewall.

   4. Leave a message for other questions, learn and discuss together

 

 

Guess you like

Origin blog.csdn.net/cvbnjmkl/article/details/108491633