[AutoQA] #396: bodhi_util: Could not connect to bodhi!
by fedora-badges
#396: bodhi_util: Could not connect to bodhi!
---------------------+------------------------
Reporter: kparal | Owner:
Type: defect | Status: new
Priority: major | Milestone: Hot issues
Component: core | Keywords:
Blocked By: | Blocking:
---------------------+------------------------
There seems to be some problem when trying to submit certain results to
Bodhi:
{{{
12/22 07:10:13 INFO | test:0478| Submitting results (log, email,
bodhi) for: depcheck: PASSED; 66 PASSED for kde-
l10n-4.7.4-1.fc16,PyKDE4-4.7.4-1.fc16,blinken-4.7.4-1.fc16,cantor-4.7.4-1.fc16,gwenview-4.7.4-1.fc16,kalgebra-4.7.4-1.fc16,kalzium-4.7.4-1.fc16,kamera-4.7.4-1.fc16,kanagram-4.7.4-1.fc16,kate-4.7.4-1.fc16,kbruch-4.7.4-1.fc16,kcolorchooser-4.7.4-1.fc16
,kde-
wallpapers-4.7.4-1.fc16,kdeaccessibility-4.7.4-1.fc16,kdeadmin-4.7.4-1.fc16,kdeartwork-4.7.4-1.fc16,kdebase-4.7.4-2.fc16
,kdebase-runtime-4.7.4-2.fc16,kdebase-
workspace-4.7.4-5.fc16,kdeedu-4.7.4-1.fc16,kdegames-4.7.4-1.fc16,kdegraphics-4.7.4-1.fc16
,kdegraphics-strigi-analyzer-4.7.4-1.fc16,kdegraphics-
thumbnailers-4.7.4-1.fc16,kdelibs-4.7.4-1.fc16,kdemultimedia-4.7.4-2.fc16,kdenetwork-4.7.4-1.fc16,kdepim-4.7.4-1.fc16
,kdepim-runtime-4.7.4-1.fc16,kdepimlibs-4.7.4-2.fc16,kdeplasma-
addons-4.7.4-1.fc16,kdesdk-4.7.4-1.fc16,kdetoys-4.7.4-1.fc16,kdeutils-4.7.4-1.fc16,kgamma-4.7.4-1.fc16,kgeography-4.7.4-1.fc16,khangman-4.7.4-1.fc16,kig-4.7.4-1.fc16,kiten-4.7.4-1.fc16,klettres-4.7.4-1.fc16,kmplot-4.7.4-1.fc16,kolourpaint-4.7.4-1.fc16,konsole-4.7.4-1.fc16
,kross-
interpreters-4.7.4-1.fc16,kruler-4.7.4-1.fc16,ksaneplugin-4.7.4-1.fc16,ksnapshot-4.7.4-1.fc16,kstars-4.7.4-1.fc16,ktouch-4.7.4-1.fc16,kturtle-4.7.4-1.fc16,kwordquiz-4.7.4-1.fc16,libkdcraw-4.7.4-1.fc16,libkdeedu-4.7.4-1.fc16,libkexiv2-4.7.4-1.fc16,libkipi-4.7.4-1.fc16,libksane-4.7.4-1.fc16,marble-4.7.4-1.fc16,okular-4.7.4-1.fc16
,oxygen-icon-
theme-4.7.4-1.fc16,parley-4.7.4-1.fc16,rocs-4.7.4-1.fc16,smokegen-4.7.4-1.fc16,smokekde-4.7.4-1.fc16,smokeqt-4.7.4-1.fc16,step-4.7.4-1.fc16,svgpart-4.7.4-1.fc16
12/22 07:10:13 INFO | test:0361| Log created:
/usr/share/autotest/results/default/depcheck/results/kde-
l10n-4.7.4-1.fc1.html
12/22 07:10:13 INFO | test:0307| Email not sent: "depcheck: PASSED;
66 PASSED for kde-
l10n-4.7.4-1.fc16,PyKDE4-4.7.4-1.fc16,blinken-4.7.4-1.fc16,cantor-4.7.4-1.fc16,gwenview-4.7.4-1.fc16,kalgebra-4.7.4-1.fc16,kalzium-4.7.4-1.fc16,kamera-4.7.4-1.fc16,kanagram-4.7.4-1.fc16,kate-4.7.4-1.fc16,kbruch-4.7.4-1.fc16,kcolorchooser-4.7.4-1.fc16
,kde-
wallpapers-4.7.4-1.fc16,kdeaccessibility-4.7.4-1.fc16,kdeadmin-4.7.4-1.fc16,kdeartwork-4.7.4-1.fc16,kdebase-4.7.4-2.fc16
,kdebase-runtime-4.7.4-2.fc16,kdebase-
workspace-4.7.4-5.fc16,kdeedu-4.7.4-1.fc16,kdegames-4.7.4-1.fc16,kdegraphics-4.7.4-1.fc16
,kdegraphics-strigi-analyzer-4.7.4-1.fc16,kdegraphics-
thumbnailers-4.7.4-1.fc16,kdelibs-4.7.4-1.fc16,kdemultimedia-4.7.4-2.fc16,kdenetwork-4.7.4-1.fc16,kdepim-4.7.4-1.fc16
,kdepim-runtime-4.7.4-1.fc16,kdepimlibs-4.7.4-2.fc16,kdeplasma-
addons-4.7.4-1.fc16,kdesdk-4.7.4-1.fc16,kdetoys-4.7.4-1.fc16,kdeutils-4.7.4-1.fc16,kgamma-4.7.4-1.fc16,kgeography-4.7.4-1.fc16,khangman-4.7.4-1.fc16,kig-4.7.4-1.fc16,kiten-4.7.4-1.fc16,klettres-4.7.4-1.fc16,kmplot-4.7.4-1.fc16,kolourpaint-4.7.4-1.fc16,konsole-4.7.4-1.fc16
,kross-
interpreters-4.7.4-1.fc16,kruler-4.7.4-1.fc16,ksaneplugin-4.7.4-1.fc16,ksnapshot-4.7.4-1.fc16,kstars-4.7.4-1.fc16,ktouch-4.7.4-1.fc16,kturtle-4.7.4-1.fc16,kwordquiz-4.7.4-1.fc16,libkdcraw-4.7.4-1.fc16,libkdeedu-4.7.4-1.fc16,libkexiv2-4.7.4-1.fc16,libkipi-4.7.4-1.fc16,libksane-4.7.4-1.fc16,marble-4.7.4-1.fc16,okular-4.7.4-1.fc16
,oxygen-icon-
theme-4.7.4-1.fc16,parley-4.7.4-1.fc16,rocs-4.7.4-1.fc16,smokegen-4.7.4-1.fc16,smokekde-4.7.4-1.fc16,smokeqt-4.7.4-1.fc16,step-4.7.4-1.fc16,svgpart-4.7.4-1.fc16"
12/22 07:10:17 ERROR|bodhi_util:0278| An error occured:
ServerError(http://aqd/bodhi/list, 500, Internal Server Error)
12/22 07:10:17 ERROR|bodhi_util:0279| Could not connect to bodhi!
12/22 07:10:17 ERROR|bodhi_util:0279| Could not post a comment to bodhi
}}}
This is running on my development machine when using mock_fedorainfra
tool. I haven't spotted it yet on either production or staging machine.
That indicates that this could be a problem inside the mock_fedorainfra
tool. I have seen this bug happening several times, always only for
updates consisting of several builds.
Investigate whether this is really a problem in mock_fedorainfra and then
act on accordingly (fix the problem/report in upstream).
--
Ticket URL: <https://fedorahosted.org/autoqa/ticket/396>
AutoQA <http://autoqa.fedorahosted.org>
Automated QA project
11 years, 10 months
Playing around with Sikuli
by Tim Flink
I've been wanting to try out Sikuli (http://www.sikuli.org) for a while
now and finally got around to making it build and work on Fedora.
The basic concept is to create GUI test cases based on images instead of
absolute (x,y) coordinates (like selenium), scripted key sequences
(openqa) or using an accessability layer (like dogtail).
I've put together a tarball with source and build scripts if anyone
else wants to try it out:
http://tflink.fedorapeople.org/tools/sikuli/README_build-sikuli.txt
http://tflink.fedorapeople.org/tools/sikuli/fedora-sikuli-20120220.tar.gz
The distribution method is a really dirty hack but I've been able to
build on both F15 and F16 after installing a huge number of -devel
dependencies and building all of Sikuli's direct dependencies from
source. Xpresser (https://wiki.ubuntu.com/Xpresser) might be another
option - it's a re-implementation of the concepts from Sikuli. It seems
to be missing some of Sikuli's features but would likely be quite a bit
easier to build and integrate.
One of the things I have in mind is to attempt automating at least some
of the installation test cases because a lot of them pass most of the
time and they can be a bit mind-numbing to do over and over again. I'm
of the opinion that our human resources could be put to better use than
monkey-button-pushing their way through simple test cases.
A non-trivial amount of work would be needed before we could fully
utilize this (fixing the code so that it is package-able, better
integration with VNC etc.) but I'm wondering if it could be a good
candidate for automating some of the installation test cases.
I've written two test cases for Sikuli thus far: a simple DVD graphical
install and graphical firstboot. I had to workaround some timing issues
due to occasional lag in the VM's VNC interface but they seem to be
working well with F17 Alpha RC2 for now.
F17 DVD Basic Graphical Install:
http://tflink.fedorapeople.org/tools/sikuli/f17-basicGraphicalDVD_sikuli....
F17 Graphical Firstboot:
http://tflink.fedorapeople.org/tools/sikuli/f17-graphicalFirstboot_sikuli...
- NOTE: This only works some of the time due to rhbz#750527 which was
closed with no bug comment by the developers. I'm going to bug them
about this but will likely update the test case to work with
unpredictable order of the firstboot screens.
To run those test cases, extract them somewhere and point the Sikuli
IDE at the extracted directories.
I'm going to be running my basic graphical install testcase for
different F17 releases to get an initial feel for how fragile the test
cases are and how much effort would be required to maintain these kinds
of test cases over a release.
Anyhow, long email over. Any thoughts on whether this might be worth
investigating farther?
Tim
11 years, 11 months
[AutoQA] #420: Some builds are tested many times over
by fedora-badges
#420: Some builds are tested many times over
---------------------+------------------------
Reporter: kparal | Owner:
Type: defect | Status: new
Priority: major | Milestone: Hot issues
Component: tests | Keywords:
Blocked By: | Blocking:
---------------------+------------------------
mkasik asked me why he has received 19 test reports from rpmlint and
rpmguard about poppler-0.18.4-3.fc18 in the last few days. It seems we
really tested the same build 19 times:
http://autoqa.fedoraproject.org/resultsdb/frontend/search?type=Envr&terms...
I looked into autoqa.log on production server, the koji-bodhi watcher
really detected it as new so many times. And it seems a lot of other
builds as well. What might be the cause?
I have looked into /var/cache/autoqa/watch-koji-
builds/koji_history_prevtimes.cache and there is no timestamp for f18
repository, which I find weird. But maybe I just don't understand the file
properly.
Anyone feels like taking this?
--
Ticket URL: <https://fedorahosted.org/autoqa/ticket/420>
AutoQA <http://autoqa.fedorahosted.org>
Automated QA project
11 years, 11 months
autoqa-stage results now kept for one month
by Kamil Paral
We run into inode issues on autoqa-stage server. I have changed the maximum time to keep old results from 2 months to 1 month. It should be enough for testing purposes.
11 years, 11 months
autoqa clients update
by Kamil Paral
I have updated all test clients associated with the staging server and also the server itself. That should resolve this ticket: https://fedorahosted.org/autoqa/ticket/415 . If you see any problems (any, not just related to that ticket), let me know. If everything looks OK, I'll update the production infrastructure tomorrow as well.
12 years
autoqa-results usage questions
by Tim Flink
I guess that the autoqa-results list is still causing some problems for
fedora-infra.
I'm honestly not sure what the problem is with our current usage of the
list - I didn't really get a clear answer to that, just that they want
to understand more about our usage in case there is a better solution
than what we're currently using.
The old archives are going to be deleted but I've downloaded a
compressed copy of the old archives in case we ever decide to parse
them.
I've been asked for the following information:
- What do we do with the list
- What do we plan to do with the list
With the recent changes that Kamil made, the only thing that I can
think of is "notification of failed or questionable test runs". Is
there anything that I'm missing?
Tim
12 years