On Mon, Dec 22, 2014 at 3:54 PM, Emmanuel Lécharny <[email protected]> wrote:
> The failing test is very time dependent. The ARM processor is most
> certainly really slow, so the test that takes around 128ms on my
> computer seems to take more than 10 seconds on your device, which is
> damn slow...

Yeah, it is a RaspberryPi, so it isn't the speediest machine around.  :)

I misunderstood the failure message, I thought it was complaining that
it completed in LESS than 10 seconds, and therefore failing.  Now that
I look at the test code, I see that you are correct.  Thanks for
pointing that out!

> You can increase the allocated time to run the test in the
> MultiThreadedMultiInvokerNotThreadsafeTest.checkRuntime() method, but I
> can assure you that even if you have this test passing, you'll face
> issues later on.

Yeah, I commented the fail() line and I'm compiling it now.

> Typically, server-integ tests are taking 3 minutes on my computer, and
> likely more than 30 minutes on ARM, which will lead Maven to stop (there
> is somehow a hard coded limit in Maven that ma    ke it 'thinks' that
> some tests lasting more than 30 minutes are a failure)...

If I run into that, I will look at cross-compiling or compiling on a
faster arm processor.  Or maybe compiling maven first with that hard
coded limit changed.

Thank you sir, for pointing out my misunderstanding!  I was
considering disabling the test, but I wanted to make sure I understood
why it was failing first.

-John

Reply via email to