Open Stack

On 11/19/2012 09:35 AM, Thierry Carrez wrote:
> Gary Kotton wrote:
>> In the past we have had specific days set aside for bug triage. How
>> about we make a concerted effort to have a test day for the following:
>>      - stable releases (if I understand correctly the up and coming one
>> is on the 22nd too)
>>      - release cycles (up and coming is the 22nd)
>> This will hopefully help up provide a more stable, robust and reliable
>> solution. In addition to this it will also enable us to identify and
>> isolate problems prior to them being found in the "real world".
>> Any ideas or comments?
>
> I think we are in two different scenarios with stable releases and
> milestone deliveries.
>
> Stable releases are directly used by our users once they are out, so if
> we want to specifically test them, it makes sense to do that before they
> are released. There usually is a call for testing in the days before the
> stable release update is issued, we could reuse that.
>
> Milestones from the current development cycle, on the other hand, are
> specifically there to tag a specific point in development to help
> testing and comparing results ("bug not present in folsom-2", "feature
> added in folsom-3"...). So we could imagine that a testing day is done
> right *after* the milestone is cut to give a common (and static) testing
> base for everyone participating.
>
> Thoughts?

Testing days are cool, and I think if someone wants to organize it, more 
power to them.

What I'd really like to see though, is a Test Writing day. Tempest's 
test set has been pretty static recently, and it would be nice to get 
more folks used to writing a test or two, especially as we're looking to 
increase the value of the test gate.

Run a test and your ensure things are stable today, write a test and you 
ensure they're stable tomorrow.

	-Sean

-- 
Sean Dague
IBM Linux Technology Center
email: sdague at linux.vnet.ibm.com
alt-email: sldague at us.ibm.com