Summary: | Response Time Vs Request and Latency Vs Request graphs don't line up with throughput | ||
---|---|---|---|
Product: | JMeter - Now in Github | Reporter: | bill.tutt |
Component: | Main | Assignee: | JMeter issues mailing list <issues> |
Status: | RESOLVED FIXED | ||
Severity: | regression | CC: | bill.tutt, p.mouawad |
Priority: | P1 | Keywords: | FixedInTrunk |
Version: | 5.0 | ||
Target Milestone: | JMETER_5.1.1 | ||
Hardware: | All | ||
OS: | All | ||
Attachments: | Load script |
Thanks for report. Indeed there is something weird on those 2 graphs. Author: pmouawad Date: Fri Feb 22 10:03:21 2019 New Revision: 1854130 URL: http://svn.apache.org/viewvc?rev=1854130&view=rev Log: Bug 63198 - Response Time Vs Request and Latency Vs Request graphs don't line up with throughput Bugzilla Id: 63198 Modified: jmeter/trunk/src/core/org/apache/jmeter/report/processor/graph/AbstractVersusRequestsGraphConsumer.java jmeter/trunk/xdocs/changes.xml Hello, Can you test the upcoming version that will be available in few minutes here: - https://builds.apache.org/job/JMeter-trunk/lastSuccessfulBuild/artifact/trunk/dist/ It should be higher to build 7099 and r1854130 Your change produces much more sensible looking output.. The dashboard total transactions/s was 22.93. The throughput Request Time Vs Request and Latency Vs Request had data points from 12 requests/s to 31 requests/s. I think the issue has been fixed. Feel free to resolve away. Thanks! This issue has been migrated to GitHub: https://github.com/apache/jmeter/issues/5018 |
Created attachment 36451 [details] Load script While trying to write up a simple validation test for JMeter for internal process reasons. A coworker noticed that the Response Time Vs Request and Latency Vs Request data didn't make any sense with JMeter 5.0/5.1 but looks perfectly ok with 4.0. As you can see the attached test is very simple and I manage to record ~1461 total requests. However, the dashboard output generated via: ...\jmeter.bat -n -t <path>\TestBing_com.jmx -l <path>\TestLogfile.data -e -o <path>\JMeterReport In 5.0 and 5.1 the total transaction per second graphed values never exceed 18. However, the Response Time Vs Request has data points at: 1000 global req/s, 2000 global req/s, 3000 global req/s, 4000 global req/s. Even the dashboard's total transactions/s doesn't exceed ~27.5. The Latency Vs Request graph has similar data points. The data in these graphs for 4.0 had more reasonable 2.0 and 7.0 global req/s data points. This seems like a bug, but is it me misunderstanding a change in 5.0 to JMeter? Thanks...