What's New in Apache JMeter 3.1

the apache jmeter team will release jmeter 3.1 in a week. the major release of jmeter 3.0 was in may and we’re already seeing new features and improvements that will improve your load testing and analysis abilities. which new abilities will you have in jmeter v. 3.1? read this blog post to find out. here are the three most important things you need to know about jmeter 3.1.

1. open-source static html report is stronger

reports and tables are crucial elements for analyzing test results. they can help you discover bottlenecks, learn about your system’s abilities, and analyze trends and statistics.

jmeter 3.1 introduces new reports and summary tables.

connect time over time graph

the connect time over time graph enables users to view and analyze how long it took each request to connect, over the different times of the test. the connection time kpi is important for learning and understanding the responsiveness of the resource tested. slow responsiveness leads to poor user experience and customer migration to competitors. knowing when connection time was high can help determine its cause and help in fixing it faster. display over time also shows trends.

response time percentiles over time (successful responses) graph

this graph displays how long it takes the system to process successful requests over time for the different percentiles of the sample size. these are important for understanding your users’ experiences while segmenting the user base. if you look at the 95th percentile, for example, you can determine the experience of the majority of users, while excluding extreme incidents. this is a better kpi for this purpose than average. as you probably know, high response time over time leads to customer frustration which results in poor business and sales. as before, the analysis over time is crucial for detecting problems and identifying trends.

source

response time overview graph

this graph presents an overview of the time it took the system to process requests, for the complete test. like the previous graphs, this display is also important for understanding the customer’s point of view of your site’s performance. if there were too many requests that had a higher response time than what you determined through your business criteria, you have a bottleneck you need to fix. the aggregation is helpful for quickly identifying system performance.

source

jmeter 3.1 will also present the top 5 errors in the sampler summary table. this identification quickly points you to the issues that need fixing. ca blazemeter also provides analysis, summary and trend reports, which you can read about here.

2. new sent bytes kpi is supported

previously, jmeter only reported received bytes. in jmeter 3.1, sent bytes are also supported. this new feature provides users with information about how response time and connect time is influenced by the number of sent bytes. this enriches your analysis abilities.

3. introducing global timer scaling factor property

pauses are an integral part of scenarios and test scripts. in real-life, customers might pause for five seconds after logging in, or pause for 10 seconds after conducting a search between items. these can be managed in load tests through jmeter timers .

however, if jmeter users want to update the test and change pause time configurations during script development, they need to manually change each timer. but not anymore. in jmeter 3.1, the global timer enables users to update all pauses in one place. this is especially important in large scripts and saves development time and meticulous work. another way to save time is the step-by-step debugging feature ca blazemeter developed and released to open-source.

these changes in jmeter, along with memory improvements and many performance improvements and optimizations, are a good reason to migrate to jmeter 3.1.

click here to learn more about performance testing with jmeter in this free webinar.