Saving Taurus report instead of Artifacts Dir

[surendra matta]

Hi Andrey,

Is there any way we can define the .JTL in Taurus config file because i need to use that .JTL in jenkins to publish the report automatically , once the Taurus complete is test or if you have any better approach  to publish the report after test completion.  pleas suggest 

Regarsd,

Venkat

[Andrey Pokhilko]

Hi,

You should override artifacts dir for Taurus for something constant like "taurus_results" and configure your jenkins to use file from that directory. Then you will get it in predictable not random location.

Andrey Pohilko Chief Scientist P: +7 (909) 631-21-69

--
You received this message because you are subscribed to the Google Groups "codename-taurus" group.
To unsubscribe from this group and stop receiving emails from it, send an email to codename-taur...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/codename-taurus/3580d62f-39e4-450d-907e-ef4b3e4474c6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[surendra matta]

HI Andrey,

Thanks for the update 

I  configure the folder directly but i am getting the below error message

00:05:30.005 Performance: Recording JMeter reports 'C:\Users\vsm754\Desktop\Jmeter-Taurus\Jenkins\Test_0229\kpi.jtl'
00:05:30.009 Performance: Parsing JMeter report file 'C:\Program Files (x86)\Jenkins\jobs\Jmeter- Taurus\builds\14\performance-reports\JMeter\kpi.jtl'.
00:05:30.015 Performance: Failed to parse file 'C:\Program Files (x86)\Jenkins\jobs\Jmeter- Taurus\builds\14\performance-reports\JMeter\kpi.jtl': For input string: "timeStamp"
00:05:30.015 java.lang.NumberFormatException: For input string: "timeStamp"
00:05:30.015 	at java.lang.NumberFormatException.forInputString(Unknown Source)
00:05:30.015 	at java.lang.Long.parseLong(Unknown Source)
00:05:30.015 	at java.lang.Long.valueOf(Unknown Source)
00:05:30.015 	at hudson.plugins.performance.JMeterCsvParser.getSample(JMeterCsvParser.java:154)
00:05:30.015 	at hudson.plugins.performance.JMeterCsvParser.parse(JMeterCsvParser.java:124)
00:05:30.015 	at hudson.plugins.performance.JMeterParser.parseCsv(JMeterParser.java:193)
00:05:30.015 	at hudson.plugins.performance.JMeterParser.parse(JMeterParser.java:51)
00:05:30.015 	at hudson.plugins.performance.AbstractParser.parse(AbstractParser.java:66)
00:05:30.015 	at hudson.plugins.performance.PerformancePublisher.perform(PerformancePublisher.java:348)
00:05:30.015 	at hudson.tasks.BuildStepMonitor$1.perform(BuildStepMonitor.java:20)
00:05:30.015 	at hudson.model.AbstractBuild$AbstractBuildExecution.perform(AbstractBuild.java:779)
00:05:30.015 	at hudson.model.AbstractBuild$AbstractBuildExecution.performAllBuildSteps(AbstractBuild.java:726)
00:05:30.015 	at hudson.model.Build$BuildExecution.post2(Build.java:185)
00:05:30.015 	at hudson.model.AbstractBuild$AbstractBuildExecution.post(AbstractBuild.java:671)
00:05:30.015 	at hudson.model.Run.execute(Run.java:1766)
00:05:30.015 	at hudson.model.FreeStyleBuild.run(FreeStyleBuild.java:43)
00:05:30.015 	at hudson.model.ResourceController.execute(ResourceController.java:98)
00:05:30.015 	at hudson.model.Executor.run(Executor.java:381)

[Andrey Pokhilko]

This is a known bug in Performance plugin. The problem is that field headers in some old Jmeter versions were "timeStamp", while latest versions have it "timestamp". This problem of uppercase "S" causes problem. It is not related to Taurus.

Here's some more context: https://issues.jenkins-ci.org/browse/JENKINS-28787

Andrey Pohilko Chief Scientist P: +7 (909) 631-21-69

To view this discussion on the web visit https://groups.google.com/d/msgid/codename-taurus/99a5cfae-a882-4a40-9f42-5b7f91eb2f80%40googlegroups.com.

[surendra matta]

Hi Andrey , 

Please let me know how to solve this issue or any other work around for this.

Regards,

Venkat

[Andrey Pokhilko]

I see no workaround here, because it's Jenkins plugin bug, the weird one.

Andrey Pohilko Chief Scientist P: +7 (909) 631-21-69

-- You received this message because you are subscribed to the Google Groups "codename-taurus" group. To unsubscribe from this group and stop receiving emails from it, send an email to codename-taur...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/codename-taurus/edd80d4f-7a7f-47a5-8717-7519adaa329d%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.

[till.n...@gmail.com]

This line:

java.lang.NumberFormatException: For input string: "timeStamp"

suggests that the plugin tries to parse the header row of the .jtl CSV file (which starts with timestamp or timeStamp, depending on Jmeter version).

Reason: The JMeterCsvParser (https://github.com/jenkinsci/performance-plugin/blob/master/src/main/java/hudson/plugins/performance/JMeterCsvParser.java#L29)

is instantiated with skipFirstLine=false (https://github.com/jenkinsci/performance-plugin/blob/master/src/main/java/hudson/plugins/performance/JMeterParser.java#L192)

Work-around: Turning off the header row apparently fixes this.

modules:

  jmeter:

    properties:

      jmeter.save.saveservice.print_field_names: false

[surendra matta]

Thanks , i will try and let you know

[till.n...@gmail.com]

Actually I just discovered that there's a separate report type JMeterCSV which you can configure to match the Taurus output (i.e. skip header row and column order).

Note that the "CSV Pattern" field only accepts certain strings which are the plugin's aliases rather than actual .jtl column names, i.e. they don't have to match the actual header row (e.g. URL vs. label, timestamp vs. timeStamp).

HTH,

Till