Open Stack

On 08/02/2014 01:05 PM, Boris Pavlovic wrote:
> Hi stackers,
> 
> 
> Could somebody explain me why we need subunit2sql? 
> 
> It seems like a part of functionality that is already implemented in Rally:
> http://www.mirantis.com/blog/rally-openstack-tempest-testing-made-simpler/
> 
> With Rally we can run tempest tests and Rally will parse & store results
> to it's DB. 
> Having one instance of Rally DB in gates will allow us to collect all
> results of all tempest runs. As well Rally is already integrated in
> gates. So there won't be issues with getting this done. 
> 
> By the way, Rally team is ready to help with this task. 
> 
> Thoughts? 
> 
> Best regards,
> Boris Pavlovic 

This is actually the crux of why I think rally belongs in the QA
program. I consider Tempest result summary to be an essential but not
yet implemented part of Tempest itself. The Rally team has implemented
some of these features we want in base Tempest, but not in a way that
can be done without using the whole Rally toolchain, which shouldn't be
required. We should be building components that can be built into the
pipeline that people want to assemble.

The Rally team is arguing that Rally is not part of QA in one thread,
then saying that all upstream QA should be done with Rally in another
thread. So now I'm even more confused by the program proposal.

	-Sean

-- 
Sean Dague
http://dague.net