Comment 6 for bug 397912

Stewart Smith wrote:
> and what IO system do these boxes have behind them? is there disk
> thrashing during the test run?
>
> other tasks running?
>
> what if setting MTR_PARALLEL=2 or MTR_PARALLEL=1 in environment (and
> checking it gets passed down okay)? Does that still have any timeouts
> reported?
>

So - I hear all of that, and that will be good in tracking it down - but
someone with a slow computer shouldn't need to tune the test suite to be
able to run it. Needing to set env variables to lower the parallel count
to run make test also unacceptable.

How about we lower mtr2's aggressiveness with bailing out due to
timeouts in general? Like, set the timeout to 15 minutes or something?

Monty