[openstack-dev] [nova] RFC - using Gerrit for Nova Blueprint review & approval
John Garbutt
john at johngarbutt.com
Mon Mar 10 12:37:57 UTC 2014
On 7 March 2014 19:50, Tim Bell <Tim.Bell at cern.ch> wrote:
>
> The recent operator gathering (https://etherpad.openstack.org/p/operators-feedback-mar14) concluded a similar proposal, based on Blueprint-on-Blueprints (BoB for short).
>
> The aim was that operators of production OpenStack clouds should engage to give input at an early stage
>
> - Raising concerns at the specification stage will be much more productive than after the code is written
> - Operators may not have the in-depth python skills for productive participation in the later stages of the review
> - Appropriate credit can also be granted to those people refining the requirements/specification process. Spotting an issue before coding starts is a major contribution.
>
> The sort of items we were worrying about were:
>
> Monitoring
> Metrics
> Alerts
> Health check/http ping
> Logging
> Lifecycle
> HA
> test
> upgrade
> roll-back
> restart
> Documentation
> object state/flow
> usage
> debug
>
> We also need to review the weighting. A blueprint should not be indefinitely postponed due to a differing view on requirements between new function and the ability to run the existing environments.
I am really keen to find a way to make this work.
I worry that getting operators to review all blueprints is not very
scalable. Although having a process where thats possible seems like a
good way to testing that out.
Getting specific pain points is clearly important. Sometimes thats
bugs, sometimes that features. But I wonder if we can agree more
general "themes" that can help inform our priority setting of bugs and
blueprints.
Back in San Deigo or San Francisco (?) I remember a good summit
session where we got lots of concrete pain points back from Ops guys
who were in the room. Making that a regular feature might be useful,
if we can use that session to tune some general priorities for the
following cycle. For example, not worried about X, what really kills
me is the lack of A and that B really sucks.
John
More information about the OpenStack-dev
mailing list