Summary: | Ability to look at summaries without GUI (non-gui) | ||
---|---|---|---|
Product: | JMeter - Now in Github | Reporter: | Martin Cleaver <martin> |
Component: | Main | Assignee: | JMeter issues mailing list <issues> |
Status: | RESOLVED DUPLICATE | ||
Severity: | enhancement | CC: | p.mouawad |
Priority: | P2 | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Hardware: | PC | ||
OS: | Mac OS X 10.4 |
Description
Martin Cleaver
2009-06-02 09:57:33 UTC
If you don't want all of the Summary Results columns, you can use: http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Generate_Summary_Results to write some info to the log file, and you can use grep or similar to extract it. Thanks, Sebb. I'd realised that. The one piece of information I need in order to mass test meaningfully is the standard deviation for any given run. Without this I can't tell whether something is awry on any given day. I'm in #jmeter on irc://irc.Freenode.net if you would like to talk. Thanks. Martin. It's unlikely to be implemented in the near future, as it would mean changing the way that Listeners work. If you use CSV output (which is more efficient), it should be fairly easy to use a stats package to extract exactly what you want, and more. Or use a scripting language such as Perl or Python, both of which have Maths libraries. Yup, doubtless I can post process. I noted that http://wiki.apache.org/jakarta-jmeter/LogAnalysis had a bunch of hacks for people wanting various data out of the system. jmeter desktop gui with functionality to calculate these figures, so duplicating the functionality in perl and other scripting languages is only going to cause duplication and rework. I realise that it is not on your roadmap and appreciate both the suggestion for workaround and listing as enhancement. Best, Martin. > jmeter desktop gui with functionality to calculate these figures, so duplicating the functionality in perl and other scripting languages is only going to cause duplication and rework. I'm wondering: would it a lot of work to factor out the analysis code such that it can work on the raw data, and not hooked in via a listener? It seems that anything that wraps around jmeter, e.g. http://wiki.apache.org/jakarta-jmeter/JMeterMavenPlugin or http://jchav.blogspot.com/ for regression testing, must be wanting this and other key outputs in its result set. I am supposing that each is duplicating the work of the stddev numbers by test type. (Though I can't see the stddev output in the JChav, perhaps it was too much work, I've just asked on the blog.) Thanks, Martin Since then, more tools have appeared, for instance http://jmeter-plugins.org/wiki/JMeterPluginsCMD/ , serving automation purposes. This issue may be closed, I suppose. Solutions exist and listed. Since Bug 58653 and with Andrei answer I think we can close this one. *** This bug has been marked as a duplicate of bug 58653 *** This issue has been migrated to GitHub: https://github.com/apache/jmeter/issues/2243 |