[Automated-testing] Test definition for CKI

[Tim.Bird at sony.com]

> -----Original Message-----
> From: Veronika Kabatova 
> 
> ----- Original Message -----
> > From: "Tim Bird" <Tim.Bird at sony.com>
> > To: vkabatov at redhat.com, nkondras at redhat.com
> > Cc: automated-testing at yoctoproject.org
> > Sent: Sunday, October 20, 2019 1:50:48 AM
> > Subject: Test definition for CKI
> >
> > Veronika or Nikolai,
> >
> > I'm putting together information for my test definition talk at ATS,
> > and I'd like to make sure I've got the information for CKI correct.
> >
> > I've been using the repository:
> > https://github.com/CKI-project/tests-beaker.git
> >
> > I seem to recall hearing about a meta-data field that referenced the source
> > files or lines,
> > that was intended to be used as part of the trigger mechanism for a test.
> > This was during your talk at Plumbers. I looked at the slides from
> > plumbers here:
> >
> https://www.linuxplumbersconf.org/event/4/contributions/287/attachment
> s/213/468/red_hat_joins_ci_party.pdf
> > and I see what I think I recall on slide 177, under pattern:sources.
> >
> > I don't see where this "Suite Data" file is or where it comes from.
> > Apparently it's in a "KPET-DB"?
> > And it looks like the file format is yaml.
> > Is there a public repository where I could look at some more examples of
> this
> > file,
> > for my test definition analysis?  Or, if not, would it be possible to send me
> > one or
> > two representative examples of this, for different tests run by CKI?
> >
> 
> Hi,
> 
> this repo is currently private as it contains some sensitive data that's way
> too mixed up with the test metadata. IIRC Nick has a plan to split it out
> and publish the repo with test data.
> 
> The full repo is used for any test run, and the kpet tool builds a Beaker
> XML based on the parameters (such as tree and architecture). The examples
> in the presentation basically contain everything about the test cases, Nick
> specifically picked the examples to show all possibilities. I don't think
> we have any docs we could point you to (definitely something we should
> improve) but if you have any specific questions that weren't answered in
> the talk let us know.
> 
> For the ease of your analysis, I put a (redacted) index.yaml for the
> networking test from the slides up here:
> 
> https://paste.fedoraproject.org/paste/YidUsdYUBt9PNOBbH15DlA

Thanks.  This is very helpful.
 -- Tim