[openstack-dev] [nova] minimum review period for functional changes that break backwards compatibility

Sean Dague sean at dague.net
Fri Jan 3 15:08:36 UTC 2014


On 01/03/2014 09:44 AM, Anne Gentle wrote:
>
>
>
> On Fri, Jan 3, 2014 at 7:52 AM, Thierry Carrez <thierry at openstack.org
> <mailto:thierry at openstack.org>> wrote:
>
>     Tim Bell wrote:
>      > Is there a mechanism to tag changes as being potentially more
>     appropriate for the more ops related profiles ? I'm thinking more
>     when someone proposes a change they suspect could have an operations
>     impact, they could highlight this as being one for particular focus.
>      >
>      > How about an OpsImpact tag ?
>
>     I think such a tag would help. That would encourage ops to start looking
>     more regularly into proposed changes by highlighting the few reviews
>     that are most likely to need their expertise.
>
>     We could have that tag post reviews to the -operators ML (in the same
>     way SecurityImpact posts to the -security ML), which would additionally
>     reinforce the need for this list as a separate list from the openstack
>     general list.
>
>
> I think this is a good idea, though I have to say when we were using
> DocImpact in this way, it kills conversational dialog on the mailing
> list because all the robots start talking on your list as more people
> use the flag. We have since moved to automatically logging a doc bug
> rather than posting to the mailing list.
>
> Still, I think it's worth trying for OpsImpact because there's not a
> secondary place to log the impactful patches. Nice idea, Tim!

If we are going down this path, it would be good to give reviewers a 
training set. So a dozen patches that someone things would be good 
OpsImpact and a dozen patches which clearly aren't.

Otherwise I fear we're making a lot of assumptions about what people 
believe are and are not OpsImpact and I don't think it will be that helpful.

I also really think that this is the less useful approach than taking 
those key ops issues into the CI toolchain. And if people only have one 
place to spend energy, spending it on the CI side is way more beneficial.

	-Sean

-- 
Sean Dague
Samsung Research America
sean at dague.net / sean.dague at samsung.com
http://dague.net



More information about the OpenStack-dev mailing list