holger at merlinux.eu
Tue Apr 28 09:05:52 PDT 2009
On Tue, Apr 28, 2009 at 09:13 -0400, Douglas Philips wrote:
> On or about 2009 Apr 28, at 6:29 AM, holger krekel indited:
> > sure, in the case of non-replicatable resources distributing
> > tests is of limited use. still you might benefit from running tests
> > in a subprocess - I consider that a form of distributing tests.
> > It allows for timeouts, and to deal with crashing processes
> > etc without disrupting the test run.
> That would be a great idea.
> Unfortunately (and this is embarrassing), we are running the tests on
> a Windows box with proprietary device drivers.(1)
> Upon loading the python shim that talks to the drivers we have to do
> device discovery to find our device(s).
> I am not savvy on the all of the "whys", but finding our device to
> test takes upwards of 20-30 seconds.
> Between the time to start a new process and the time to (re-re-re-re-
> re-)discover the testing device, our testing times would explode if we
> ran each test in a subprocess. :(
maybe i am missing something, but i am not talking about
isolating each test in a fresh subprocess, rather a mode
where you run multiple tests in one process and restart
the process in case one test crashed it. This way
you would look for device drivers N+1 times with N
being the number of crashes of your tests (and/or the
number of timeouts etc).
> (1) - testing is almost always more "interesting" that correct usage,
> so we need to have custom drivers to allow us access that Windows
> doesn't normally permit. Further, we are using Windows. That is
> something I probably can't change :( :( :( - there are only so many
> aircraft carriers to steer, and we're lucky to be using Python. But
> those battles are really off-topic here.
> testing-in-python mailing list
> testing-in-python at lists.idyll.org
Metaprogramming, Python, Testing: http://tetamap.wordpress.com
Python, PyPy, pytest contracting: http://merlinux.eu
More information about the testing-in-python