Hi Ebenezer,

thanks for the good feedback and proposing enhancements.

If you use pull requests then (normally) only your own build can fail, you
shouldn't be able to merge any pull request that fails.
So the information that build failed is relevant only for the authors of
the pull request not all the developers of Fineract.
Develop branch should normally never fail (as long as anyone doesn't bypass
submitting the pull request).

Regarding the logs - you can get the build logs directly from travis via
web.
For example - there is a failing pull request here at the moment:
https://github.com/apache/fineract-cn-demo-server/pull/27
If you expand Travis and then click
 The build
<https://travis-ci.com/apache/fineract-cn-demo-server/builds/111226324>
failed
then you get the full console output of the failing build.

I hope I understood your poroblem right but feel free to elaborate.

Kind regards
Juhan



Kontakt Ebenezer Graham (<egraha...@alustudent.com>) kirjutas kuupäeval N,
23. mai 2019 kell 23:41:

> Hi Community / Juhan,
>
> Nice job with the Travis CI integration. They are working nicely! In
> addition, I want to suggest an idea.
>
> Can we have an after failure procedure in the .travis.yml to make the log
> output of a failed build available to Fineract developers?
>
> As you may already know, with TravisCI, there is no way to login/ssh into
> a build machine to inspect the job's log output interactively. And so for
> instance, zipping and publishing rat reports, component-test and service
> build logs for analysis will come in handy for a number of developers as we
> venture into CI/CD.
>
> I believe the common approach is by defining an after_failure script to
> dump the logs in to log aggregator of our choosing. The major problem here
> will be "where to publish the logs"
>
> Also, I will create an issue on Jira to keep track of this suggestion.
>
> Looking forward to your response.
>
>
> *At your service,*
>
> *Ebenezer Graham*
>
>
>
>
> On Thu, 25 Apr 2019 at 15:43, Juhan Aasaru <aas...@gmail.com> wrote:
>
>> Hey,
>>
>> thank you for feedback and help Vishwas and Courage.
>> I have now created 30 pull requests for getting travis to build and
>> upload to artifactory, all linked here ander Issue links:
>> https://issues.apache.org/jira/browse/FINCN-148
>> Its mostly copy-and-paste. Travis now builds all pull requests including
>> the requests created by me.
>>
>> For identity I also changed the Dockerfile to wget the artifact from
>> Artifcatory.
>> I didn't add this to any other project yet (I plan to keep it a separate
>> task).
>>
>> I hope some reviewer has a chance to go over my pull requests.
>>
>> Kind regards
>> Juhan
>>
>>
>>
>> Kontakt Courage Angeh (<couragean...@gmail.com>) kirjutas kuupäeval R,
>> 12. aprill 2019 kell 02:27:
>>
>>> Hi Juhan,
>>>
>>> Nice work!. The code base and documentation looks good.
>>>
>>> Regards,
>>> Courage.
>>>
>>> On Thu, Apr 11, 2019 at 7:16 PM Vishwas Babu <
>>> vish...@confluxtechnologies.com> wrote:
>>>
>>> > +1
>>> > Looks good to me
>>> >
>>> >
>>> > Regards,
>>> > Vishwas
>>> >
>>> >
>>> >
>>> > On Thu, Apr 11, 2019 at 9:13 AM Juhan Aasaru <aas...@gmail.com> wrote:
>>> >
>>> > > Hi!
>>> > >
>>> > > I have proof-of-concept ready and I have managed to build and publish
>>> > > fineract-cn-identity artifacts with Travis.
>>> > > All the fineract-cn-* projects that idenity depends on I made forks
>>> and
>>> > set
>>> > > Travis to build and publish those artifacts.
>>> > > Before I go over rest of the projects and create pull requests please
>>> > have
>>> > > a look and comment if you have time
>>> > > (I think most important is to agree on artifact naming for
>>> non-develop
>>> > > brannches and review changes to *.gradle files)
>>> > >
>>> > > The idea is that Travis builds snapshot artifacts and uploads them to
>>> > > https://mifos.jfrog.io/mifos/libs-snapshot/org/apache/fineract/cn/
>>> > > (I created user "travis-ci" to artifactory with upload permissions
>>> and
>>> > > encrypted its password for each repository)
>>> > > Travis detects branch automatically and if branch=develop (in my code
>>> > > currently i have set branch=travis) then it uses version set in
>>> > > build.gradle
>>> > > For any other branch it sets artifact version to branchname-SNAPSHOT
>>> > > If one creates a release in github then it uses tag as version and
>>> > uploads
>>> > > it under different repoKey,
>>> > > so that it will be available here:
>>> > > https://mifos.jfrog.io/mifos/libs-release/org/apache/fineract/cn/
>>> > > Travis will also build all pull requests but those artifacts are not
>>> > > uploaded anywhere.
>>> > >
>>> > > I have documented the usage and setup here:
>>> > >
>>> > >
>>> >
>>> https://cwiki.apache.org/confluence/display/FINERACT/Fineract-CN+Artifactory
>>> > >
>>> > > Here is travis build log for Idenity project:
>>> > > https://travis-ci.com/aasaru/fineract-cn-identity
>>> > >
>>> > > And the code changes of fineract-cn-idenity:
>>> > >
>>> > >
>>> >
>>> https://github.com/apache/fineract-cn-identity/compare/develop...aasaru:travis
>>> > > All the other repos github.com/aasaru/fineract-cn-* will have
>>> similar
>>> > > changes
>>> > > (travis.sh will be identical everywhere and .travis.yml as well
>>> (except
>>> > the
>>> > > encrypted password)
>>> > >
>>> > > Please let me know if you have any comments or if you have none then
>>> I
>>> > will
>>> > > create pull requests next week.
>>> > >
>>> > > Kind regards
>>> > > Juhan
>>> > >
>>> > >
>>> > > Kontakt Isaac Kamga (<isaac.ka...@mifos.org>) kirjutas kuupäeval P,
>>> 31.
>>> > > märts 2019 kell 18:59:
>>> > >
>>> > > > Hello Juhan,
>>> > > >
>>> > > > Yes, get started with the proof of concept. You'll also need to
>>> create
>>> > > > issues for the CI/CD work you're doing and probably name branches
>>> after
>>> > > the
>>> > > > issue number.
>>> > > >
>>> > > > Vishwas mentioned that a Jfrog open source license is available for
>>> > > Apache
>>> > > > Fineract. I think he's best placed to get those details though to
>>> you.
>>> > > >
>>> > > > More grease to your elbows.
>>> > > >
>>> > > > Cheers,
>>> > > > Isaac Kamga.
>>> > > >
>>> > >
>>> >
>>>
>>

Reply via email to