Al Stone wrote:> Based on some discussion held during the Xen Summit, I have started
> trying to organize the work that people are doing testing Xen. Granted,
> this bears a remarkable similarity to herding cats.
>
> Nonetheless, I''ve started collecting information about all of the
> tests that people run (many of them regularly) on the Xen wiki:
>
> http://wiki.xensource.com/xenwiki/XenTest
Excellent start, thanks very much, Al!
BTW, I was still waiting on Xen Summit presentations
to go up on the XenSource wiki. Anyone have any news
on that?
> What I propose it that we as a community decide on what it means
> when we say ''Xen works correctly'' -- either for a release
or a new
> platform. That is to say, what is the minimal set of functionality
> required to consider a port ''working''? And how do we
determine that
> that functionality _is_ working? Ideally, we can build on the most
> excellent work already done in xm-test and maybe even someday get
> to the point where a developer can run ''make check'' and
know whether
> or not their patch helped or hindered.
Agreed.
We''re trying very hard to have a simple and functional
testframework and testsuite that _anyone_ can run,
including, and perhaps especially developers.
A common framework that we can add tests to would be
invaluable in leveraging the efforts of others here.
Would be good to get most of the effort in the community
put into writing test cases as opposed to duplicate
infrastructure work...
> Any help that you can offer -- corrections, wiki editing, better
> testing methodology -- is greatly appreciated. Hit the wiki site
> and let''s see if we can decide something by the _next_ Xen Summit;
> what''s there is a very rough first pass and I''ll be the
first to
> admit it still needs a _lot_ of work :).
Yep, will do.
thanks,
Nivedita
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel