AFAIR The test was stable before migration to gradle.

Maybe more parallelism in test lead to such instabilities.

I don’t think it should be disabled

Regards

On Monday, August 12, 2019, sebb <[email protected]> wrote:

> On Mon, 12 Aug 2019 at 21:09, Vladimir Sitnikov
> <[email protected]> wrote:
> >
> > TL;DR:
> >
> > org.apache.jmeter.samplers.TestSampleResult > testSubResultsFalsePause
> FAILED
> >     java.lang.AssertionError: TestElapsed: 424 -  ParentElapsed: 401
> > => 23 not in [0,16]; nanotime=false
> >         at org.junit.Assert.fail(Assert.java:88)
> >         at org.apache.jmeter.samplers.TestSampleResult.testSubResults(
> TestSampleResult.java:292)
> >         at org.apache.jmeter.samplers.TestSampleResult.testSubResults(
> TestSampleResult.java:198)
> >         at org.apache.jmeter.samplers.TestSampleResult.
> testSubResultsFalsePause(TestSampleResult.java:181)
> >
> >
> >
> >
> > Should we disable the test?
>
> Only if it cannot be fixed or it serves no purpose.
>
> > Can we write the test in such a way that it does not break half of the
> > times in CI?
>
> Increase the time limit?
>
> > Is the test doing any good for us?
> >
> > It is really annoying to have flaky tests for little reason.
>
> Timing tests are always tricky to get right, but that does not mean
> they are not useful.
>
> Maybe it can be reworked so that it allows for a slower system.
>
> > Vladimir
> >
> > пн, 12 авг. 2019 г. в 23:02, Apache Jenkins Server <
> > [email protected]>:
> >
> > > See <https://builds.apache.org/job/JMeter-trunk/7296/display/redirect>
> > >
> > >
>

Reply via email to