Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

Summary of error reporting Solutions for Jmeter distributed execution

2025-01-29 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Internet Technology >

Share

Shulou(Shulou.com)06/03 Report--

When you use JMeter to perform distributed testing, you often encounter all kinds of errors. This paper summarizes several problems that you often encounter in your work.

1. The master side of this article is windows, and the slave side is linux.

2. No matter whether master and slave are in windows or linux, first of all, make sure that the network is connected, and the telnet ip port is fine.

3. It is suitable for the platform to call jmeter in large quantities to implement performance testing and troubleshooting. The port is occupied and the slave is abnormally shut down, resulting in the zombie process on the master side.

1. Perform performance stress test and log prompt Failed to create UDP port (failed to create UDP port)

This problem occurs because the slave process listening port of jmeter is occupied, so that the port cannot be used. The error is as follows

To solve the problem, you only need to modify the jmeter.property file of the master machine. The configuration file is located in the bin directory of jmeter. The default configuration only provides a range of 10 ports for listening, and you only need to modify it:

The   can be modified as follows

Second, perform the performance stress test. The abnormal interface for the execution part of the master log is as follows. The execution process of the task only shows part of the execution result and will not be completed successfully.

The most likely reason for this problem is that the slave process has been accidentally shut down, such as insufficient memory, resulting in an abnormal shutdown, and the master cannot listen for slave messages, resulting in hanging all the time.

First, use netstat-anp | grep 1100 on the slave machine to check whether the port occupied by slave is normal. The following figure indicates that the port is not occupied, and the slave process has been closed, so the master has been stuck, so you need to restart slave and re-stress test. If a memory overflow is reported in the jmeter-server.log log of the slave machine, you need to modify the default maximum memory.

When this problem occurs, the master process will not shut down and will always occupy the system resources. Windows can directly close the execution interface, and the linux page execution needs to kill the master process.

3. Perform performance stress test. Master log does not return any execution result. Slave has execution log.

The exception interface is as follows. The execution process of the task only shows part of the execution result and will not be completed successfully.

When this problem occurs, first check the jmeter-server.log log of the slave machine. In the bin directory, you can see the exception prompted by the log to connect 169.254.217.204. Check the local ip and find that the ip is a virtual network card. You only need to disable the network connection.

When the distributed test is executed again, the master execution log is displayed normally and the report is generated normally.

First of all, to ensure that the network of master machines and slave machines is connected, you need to execute telnet 10.16.69.174 1100 on master. Ip is the ip of the slave machine, and the port is the port set on the slave side. If you can enter the command interface normally, it is proved that the connection to the slave end is connected.

4. Perform performance stress test. Log prompt Engine is busy-please try later error reporting interface is as follows

The problem occurs, that is, the slave is occupied and not released, and the master execution may not be closed normally, resulting in the slave being occupied all the time. You only need to shut down the slave service and restart it. The operation command is as follows: first, netstat-anp | grep 1100, query the slave process id, and then kill-9 id.

    if the article is helpful to you, you are welcome to follow my official account, which is synchronized with the articles on this platform for easy reference. I will continue to publish articles related to testing and share testing techniques with you. Every original article is written attentively to put an end to plagiarism.

QQ technology exchange group: please enter the verification information 51CTO to add the group.

Wechat QR code follow official account:

After following, reply to the resource download, and you can get the download address of all kinds of resources shared by yourself.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 250

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

Internet Technology

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report