[openstack-dev] Nova PTL candidate questionnaire.

Russell Bryant rbryant at redhat.com
Thu Mar 7 17:50:26 UTC 2013


On 03/07/2013 01:41 AM, Joshua Harlow wrote:
> Just some initial questions that might help me formulate who is a good
> candidate (in my opinion of good).

Hi!

Since it contains relevant information and detail, here is a link to my
original message for reference:

http://lists.openstack.org/pipermail/openstack-dev/2013-March/006222.html

>  1. What do you think during your term are 3 things that you would like
>     to make better in nova?

I listed three things in my candidacy email.  Summary:

a) Nova project structure and leadership: encourage sub-teams,

b) Collaboration with other projects, and in particular, Quantum.

c) Some technical things: security (trusted messaging and then access
control built on it), and scalability (cells, among other things)

>       * How will you use customer feedback (people deploying openstack
>         and/or users using openstack) to influence those decisions? 

That depends on what the feedback is, I guess.  :-)

This feedback is certainly important to me.  (See my answer to the last
question, which is somewhat related to this)

>  2. What are 3 things that you believe nova is doing right and wrong?
>       * How would you make those 3 things better?

a) I think we're doing reviews right and wrong.  We're hugely productive
(over 11,000 nova+novaclient reviews in the last 6 months!).  We could
stand to be even more productive as a team judging by the review queue
at times.  I think there is room for improvement on the consistency and
depth of our reviews, as well.

Improving this includes ensuring that we have the right people on the
review team (nova-core), and that all of those people are engaged and
communicating well with each other.  It also includes making sure that
we have appropriate metrics available to see how well we're doing with
keeping up.

b) I think we're doing bugs right and wrong.   We did a much better job
staying on top of triage this cycle, so that's good.  We could do a much
better job fixing bugs reported by others, especially outside of the
feature freeze.

Again, I think having the right metrics available to track our status
and progress can help.  I also think that generally stressing the
importance of bug handling as a strategic contribution to OpenStack is
important.  We need companies to direct resources at it as an important
area for the success of the project.

c) Documentation is another thing that we do both right and wrong.
Documentation comes in many forms.  I think we have drastically improved
our documentation of individual changes in commit messages.  We've been
more consistent with the usage of the DocImpact keyword to flag
documentation impacting changes.

We could do a much better job in general of documenting these features
ourselves instead of often relying on doc writers to figure it out
themselves later on from the commit message, code, trial and error, or
whatever else.  This can be improved by strongly encouraging feature
documentation as a part of the code review process.

We could also do a much better job with design documentation.  Some
blueprints have excellent documentation associated with them.  Others
(even some of my own, admittedly!) just include a brief description and
nothing else besides the actual implementation.  We can improve this as
a part of the blueprint review process.


>  3. Nova adds a lot of new features each release, do you think that is
>     good, bad, or in between? 

Good.  The project is not old and we have many gaps still to fill.  It's
an expected result of the stage of the project we are in, IMO.

>       * What are your thoughts on new features vs. stability and how
>         will you address that during your term? Do you believe this is a
>         problem to begin with?

I think we have a good release process that aggressively ramps down what
can go in toward the end of the release cycle.  I also think the stable
maintenance team has a good set of policies for what is allowed in
stable branches.

As mentioned above, we could certainly use more consistent focus on bug
reports that come in, but in general I do not think we're in bad shape
here.  Our CI efforts really help this.

>  4. Scale is a big question that is always interesting, how have you
>     used openstack at scale, and what lessons have you learned that can
>     make nova better during your term

I haven't used OpenStack at scale, personally.  My career has been
focused on open source software development, often including leadership
roles.  A huge part of doing that successfully is talking to people that
use the software.

In my job, I interact with people doing deployments internally.  I also
interact with some customers, including ones doing very large deployments.

I'm on IRC, mailing lists, and attend various conferences.  All of these
are forums that I use to find feedback from users of the software that I
work on.  I usually end up identifying a number of people that
consistently provide excellent feedback and pay even closer attention
when they speak up.  This has been pretty effective for me.

Thanks,

-- 
Russell Bryant



More information about the OpenStack-dev mailing list