jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jukka Zitting" <jukka.zitt...@gmail.com>
Subject Re: svn commit: r633471 - in /jackrabbit/trunk/jackrabbit-core/src/test/java/org/apache/jackrabbit/core: ./ integration/ query/
Date Tue, 04 Mar 2008 14:29:40 GMT

On Tue, Mar 4, 2008 at 4:11 PM, Marcel Reutegger
<marcel.reutegger@gmx.net> wrote:
> Jukka Zitting wrote:
>  > It should be up to each test case to know how long it needs to run to
>  > verify some behaviour.
>  that means we will have a much longer 'mvn install'. if that's OK with everyone
>  then we don't need the scale variable and I'll remove it again.

My point is that if a test case requires more than a few seconds for a
reasonable run, then we should not compromise that requirement just to
make it a part of the standard build.

What's the use in running a test case if we know it won't fail even if
the related fix isn't there? And on the other hand, if a test produces
reasonably repeatable results at scale X, what's the point in running
it at larger scales?

On Tue, Mar 4, 2008 at 4:20 PM, Esteban Franqueiro
<esteban.franqueiro@bea.com> wrote:
>  What about having a separate test suite that doesn't run with 'mvn install'?
>  Since this tests are not unit tests, we can isolate them and run them with
>  'mvn -Dtest=PerformanceTests' or something similar.

+1 We can also set up the CI server to run such longer test cases for
example daily to catch any regressions.


Jukka Zitting

View raw message