[Qgis-developer] PyQgsServerWFST disabled on Travis

[Alessandro Pasotti]

On Mon, Oct 3, 2016 at 12:11 PM, Matthias Kuhn <matthias at opengis.ch> wrote:

>
>
> On 10/03/2016 12:04 PM, Alessandro Pasotti wrote:
> > On Mon, Oct 3, 2016 at 11:28 AM, Nyall Dawson <nyall.dawson at gmail.com
> > <mailto:nyall.dawson at gmail.com>> wrote:
> >
> >     On 3 October 2016 at 19:11, Alessandro Pasotti <apasotti at gmail.com
> >     <mailto:apasotti at gmail.com>> wrote:
> >     > On Mon, Oct 3, 2016 at 10:44 AM, Matthias Kuhn <
> matthias at opengis.ch <mailto:matthias at opengis.ch>> wrote:
> >     >>
> >     >> Following the conversation from
> >     >> https://github.com/qgis/QGIS/commit/89d4998
> >     <https://github.com/qgis/QGIS/commit/89d4998> , this is an
> announcement
> >     >> to notify that the PyQgsServerWFST has been disabled on the
> Travis builds.
> >     >>
> >     >> Work has been underway to remove "noisy" tests from Travis which
> cause
> >     >> false alarms and these tests regularly fail for unrelated changes.
> >     >> False alarms are bad as they undermine confidence in the CI
> results
> >     >> and encourage contributors to ignore the test results.
> >     >>
> >     >> If you're interested in seeing these tests continue as part of
> the CI
> >     >> infrastructure here are some examples of typical random failures
> these
> >     >> tests encounter:
> >     >>
> >     >> PyQgsServerWFST: https://travis-ci.org/qgis/QGIS/builds/164559513
> >     <https://travis-ci.org/qgis/QGIS/builds/164559513>
> >     >>
> >     >> Matthias
> >     >
> >     >
> >     >
> >     > Hi Matthias,
> >     >
> >     > Is there any way to see the complete build logs for that
> particular build?
> >     >
> >     > I'm trying with dash but the interesting part was apparently cut.
> >     >
> >     > The rest of the test output was removed since it exceeds the
> threshold of
> >     > 307200 bytes.
> >
> >     If you make the very first output from the test a 'CTEST_FULL_OUTPUT'
> >     line, then the truncation doesn't occur.
> >
> >     test_qgsserver_accesscontrol.py uses this, have a look how it's done
> >     there.
> >
> >     Nyall
> >
> >
> >
> > Thanks! I will do it.
> >
> > BTW, restarting the build
> > https://travis-ci.org/qgis/QGIS/builds/164559513, the WFS-T test now
> > passes but there is another one that fails:
> > https://travis-ci.org/qgis/QGIS/builds/164559513
> >
> > PyQgsComposerPicture ....................***Timeout 60.00 sec
> >
> > Should we disable that one too?
>
> It's already disabled, the regex for blacklisting was broken and it was
> executed by mistake.
> Fixed in 8e54acf
>
> Matthias
>
>
Hi,

I cannot see any other reason for the random failures than the "sleep(x)"
that the test waited for the server to accept connections.

I've changed the check to a more robust wait-until-ready-or-timeout
function, this should eliminate those odd sparse failures.


Please let me know if you still have problems with these tests.

https://github.com/qgis/QGIS/pull/3563


-- 
Alessandro Pasotti
w3:   www.itopen.it
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-developer/attachments/20161003/8e534816/attachment-0001.html>