activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gary Tully <>
Subject Re: Moving slow tests into a build profile.
Date Thu, 26 Jul 2012 10:06:40 GMT
I think this is really beneficial for dev and a very positive
improvement. Disabling 402 is fine.

We need to ensure that the default builds for ci/jenkins is to do the
complete run though. Having this run nightly is still important.

There are some scenarios where you have to pump message volume into
the system and most of those long running tests are the result of real
life use cases reduced to repeatable tests.
My feeling is that the work required to reduce them to something
shorter would be a little risky from a regression perspective.

On 26 July 2012 03:57, Hiram Chirino <> wrote:
> Hi folks,
> I don't know about you guys but I'd rather have the default developer build
> finish in some sort of reasonable time.
> I took the build output of one of CI builds and figured out how much time
> is spent on running the tests.  It came out to 8.03 hours.  Then sorted all
> the tests and tried to figure out how many tests I would need to disable to
> spend at most 30 min.  These are my results:
> If you sort by total test time, then you only have to disable 263 of the
> test classes and it ends up running 1824 test methods.  If you sort by
> 'test time/test methods' then you have to disable 402 test classes but it
> ends up running 3026 test methods out of a total of 4777 methods.
> I would like to move the 'slow' tests off to it's own builds profile.  The
> question is which tests do I pick to disable. Should I disable the 263 or
> the 402.  I'm leaning towards the 402 since you end up running more test
> methods.
> Once we do have the slow list checked into a pom, perhaps folks can try to
> optimize those test so that we can bring them back to the standard build.
>  WDYT?
> --
> **
> *Hiram Chirino*
> *Software Fellow | FuseSource Corp.*
> * |*
> *skype: hiramchirino | twitter: @hiramchirino<>
> *
> *blog: Hiram Chirino's Bit Mojo <>*
> *
> *
> *
> *


View raw message