Apache Software Foundation (ASF) is a non-profit charitable organization headquartered in the United States. All ASF products are developed through collaborative online public forum, and distributed from a central server in the United States. Jmeter ASF is the open-source freeware, in many domestic small and medium companies have been widely used as performance testing tool. But a lot of friends in the network feedback Jmeter of inaccurate test results, let's talk about the main reason inaccurate.
Jmeter do performance or stress testing, we must first establish a thread group, if you need to add a transaction in the transaction thread group, then you can set up a virtual number of concurrent users as necessary, also set the number of threads running concurrently.
In order to understand jmeter statistical results are not accurate, we need to find out the mechanism of the thread group Jmeter HTTP requests to run. For Jmeter, HTTP requests within the linear execution thread group (after a request to execute and return results on, then the next request requests), but in fact is often not so simple. Chrome follows access a web page, HTTP requests are concurrently carried out (previous request has not yet returned the result, the next request is issued to start request). This will lead to Jmeter its excellent statistical services is a greater deviation.
