[openstack-dev] [Metrics] Keeping up with review requests

Sean Dague sean at dague.net
Tue May 28 17:18:52 UTC 2013


On 05/28/2013 12:46 PM, Stefano Maffulli wrote:
> On 05/28/2013 07:30 AM, Russell Bryant wrote:
>> We currently have over 150 review requests for the master branch of
>> nova.  I have also been pinged 5 times this morning from people asking
>> for attention on their review request.  :-)  This makes me feel behind,
>> but that may not really be the case.  If everything has been recently
>> updated, it's just an incredibly active project.  If a large number of
>> them have been waiting a long time, we're behind.  It's important to
>> know what the real status is.
>
> Thank you for the feedback, this is really helpful to drive the Activity
> Board forward. You're not the first one that asked about reports for
> quantity and speed of reviews. At the moment there are a few pages that
> may help you get an overview of what's going on for Nova and the other
> projects. For example on
>
> <http://activity.openstack.org/data/display/OPNSTK2/OpenStack+Compute+%28Nova%29+-+Activity+Dashboard>
>
>
> you'll see the chart with the total number of reviews in different
> states per day. I've started discussing with Bitergia and zAgile how to
> add the votes and comments and the time for a change to go through to
> the reports. I hope to have more details for you soon.
>
> The Nova page has the list of bugs that have seen activity in the past
> 30 days
> <http://activity.openstack.org/data/pages/viewpage.action?pageId=3178511
> (and we're adding the list of changes/reviews soon).
>
> You can see the list of people active on the project with links to their
> personal pages on:
>
> <http://activity.openstack.org/data/display/OPNSTK2/OpenStack+Compute+%28Nova%29+-+Contributors+Activity>
>
>
> For example, Russel:
>
> <http://activity.openstack.org/data/pages/viewpage.action?pageId=3670022>

Honestly, I think this misses the entire point. These are really more 
"rah rah" kinds of stats, and not the kind of useful reports that a core 
reviewer wants out of the gerrit data.

Whatever the answer is, it should be written in a language that the PTLs 
are fluent in, because they are definitely going to want to tweak these 
algorithms over time. And if it's more than one person's private 
scripts, it should be something that's up on 
http://github.com/openstack-infra so that it's subject to the same kind 
gerrit review flow and collaboration.

	-Sean

-- 
Sean Dague
http://dague.net



More information about the OpenStack-dev mailing list