canonical-ci-engineering team mailing list archive

[Brendan Donegan <brendan.donegan@xxxxxxxxxxxxx>]

Ugh, sigh - EOD emails. Of course there were supposed to be links (for
merge requests, no bugs were ever raised for these issues, myself and ogra
just went and fixed them):

[1]
https://code.launchpad.net/~ogra/phablet-tools/phablet-tools-fix-writable-image/+merge/234946
,
https://code.launchpad.net/~brendan-donegan/phablet-tools/phablet_click_test_setup_ppa/+merge/235007

Indeed - it would be good if this was integrated with the CI test run for
phablet-tools so that changes to it would not be landed if they broke
stuff. However, that would't have caught the first of these issues so we do
need a job that runs as a smoketest on the image too.

On Wed, Sep 17, 2014 at 11:42 PM, Thomi Richards <
thomi.richards@xxxxxxxxxxxxx> wrote:

> Hi,
>
> comments below...
>
> On Thu, Sep 18, 2014 at 4:18 AM, Brendan Donegan <
> brendan.donegan@xxxxxxxxxxxxx> wrote:
>
>> A couple of recent incidents with phablet-tools [1] have highlighted that
>> we don't seem to have any smoketesting jobs that run with a silo installed.
>> This meant that we went several days, maybe even weeks with some things
>> being really broken and we don't know what the impact was. In the case of
>> the first bug listed below it's not beyond the realms of possibility that
>> people were not testing Utopic silos *at all* - which is a pretty bad
>> situation to be in.
>>
>>
> I'm missing a bit of context here... I think you forgot to add the [1]
> footnote, and there's no bug listed that I can see :D
>
>
>> Paul had a concern about bringing up these jobs and then nobody looking
>> at the results, so I volunteer to receive notifications from Jenkins when
>> these jobs fail. If we agree on this I can work with Paul to bring up the
>> jobs.
>>
>
> It sounds like this would be a jenkins job that tested phablet-tools
> itself? If my understanding is correct, then I'd like to know why this
> can't be part of their regular CI test runs (so code to phablet-tools can't
> land if the tests fail). Is it simply that the tests require real hardware?
>
> It sounds like a good idea (more testing! yay!), I just need some more
> specific information.
>
>
> Cheers!
>
> --
> Thomi Richards
> thomi.richards@xxxxxxxxxxxxx
>