[almighty] [Planner] demo.almighty Dropout Yesterday

[Max Andersen]

Great you resolved the issue Michael!

One thought about build being published when failing. Maybe it has same issue as devdoc had and script doesn't return the right return code. 



/max
http://about.me/maxandersen


> On 13 Jan 2017, at 09:58, Michael Kleinhenz <kleinhenz at redhat.com> wrote:
> 
> Hi Team,
> 
> the demo.almighty.io service went down yesterday for a couple of
> minutes. The problem is now fixed and the server is back up, but I
> found two underlying, more general problems. Here's my findings in
> bullet points:
> 
> - We had a problem on code level yesterday in master. That problem
> should have been easily
>  detected by our test chain. But it was not.
> 
> - currently, the functional tests are run against an intermediate
> build. We're using the
>  WebPack dev/debug function for that. Only after they succeed, the
> final build is done
>  and the output artifact created. The reason is that we don't need a
> test-only http-server setup
>  as WebPack already includes that.
> 
> - I found out that the error caused by the problem in master was
> apperaring ONLY when
>  doing a production build, NOT when doing a debug/dev build. The
> result was, that the problem
>  was completely invisible in a dev environment and for the functional
> tests. I don't know yet, why
>  the dev and prod build done by WebPack are behaving differently at a
> such elementary level.
> 
> - The above WebPack mystery would have not affected demo.almighty
> without a second problem:
>  the build was deployed to demo.almighty althought it failed
> building. Resulting in just index.html
>  deployed to the server (!).
> 
> - I suspect that errors appearing inside the build container are
> somehow not propagated to the
>  outside Jenkins build. They may just be consumed somewhere.
> 
> - The second problem (build errors do not get propagated) is the more
> severe one and also may
>  affect other builds in other projects as we basically just re-used that setup.
> 
> Action points:
> 
> - We need to check if we can run the functional tests against the
> production build instead of the
>  intermediate build.
>  -> https://github.com/fabric8io/fabric8-planner/issues/741
> 
> - We need to check the build process to find out why errors are not
> propagated to the outside
>  world (or being recognized as a fail by Jenkins).
>  -> https://github.com/fabric8io/fabric8-planner/issues/742
> 
> - We need to fix the actual code problems that triggered the above.
>  -> https://github.com/fabric8io/fabric8-planner/issues/732
> 
> -- Michael
> 
> -- 
> Michael Kleinhenz
> Principal Software Engineer
> 
> Red Hat Deutschland GmbH
> Werner-von-Siemens-Ring 14
> 85630 Grasbrunn
> Germany
> 
> RED HAT | TRIED. TESTED. TRUSTED.
> Red Hat GmbH, www.de.redhat.com,
> Registered seat: Grasbrunn, Commercial register: Amtsgericht München,
> HRB 153243,
> Managing Directors: Paul Argiry, Charles Cachera, Michael Cunningham,
> Michael O'Neill
> 
> _______________________________________________
> almighty-public mailing list
> almighty-public at redhat.com
> https://www.redhat.com/mailman/listinfo/almighty-public
> 
>