7fb5c479dd
This is my official candidacy for the role of Glance PTL for Newton. Change-Id: If750e59ee1c843aa88396f230f5b8afc24dadc42
85 lines
5.0 KiB
Plaintext
85 lines
5.0 KiB
Plaintext
Hello everyone,
|
|
|
|
I would like to announce my candidacy for the role of PTL of Glance for Newton.
|
|
|
|
1. Personal Commitment
|
|
My employer will be giving me 100% time to focus on this role and I do not
|
|
intend to take other roles within the community for the sake of focusing on the
|
|
development and progress of our project. I am as much of a community developer
|
|
as I like being a software one and of late, I've been focusing hard on the
|
|
first more. Thanks to this community I was given the challenging yet
|
|
interesting task of leading Glance through a difficult phase of an identity
|
|
crisis during Kilo timeframe. We've come quite far, addressed many concerns and
|
|
are making good progress therein. However, there's a lot ahead of us like the
|
|
recent image locations usage confusion, where tasks are heading and what do
|
|
image sharing plans look like. This keeps me driven and interested in the
|
|
development of Glance.
|
|
|
|
2. Themed Focus
|
|
For newton, we already have over a dozen solid proposals for summit sessions
|
|
[1] and I expect a few more. While we welcome all the feedback that we can
|
|
possibly get in such face to face conversations, it will be essential deriving
|
|
priorities out of them. Without a doubt, we've import refactor work, Nova's
|
|
support of v2 API as top priorities. Those will be the encompassing
|
|
development goals for Newton. However, any software project has some general
|
|
changes that sometimes become critical if not fixed in a limited time frame and
|
|
we've seen that happen multiple times over the history of the project. I would
|
|
like to divide those into the following focus areas: security, stability,
|
|
feature additions, performance, maintenance, testing, usability, and
|
|
cross-project. Not all the dimensions of the project will have proposal talks,
|
|
we will still encounter refactors and bugs for each. Sometimes those bugs are
|
|
heavy to handle with multiple hundred lines of code per review. Historically
|
|
Glance core being a overall low bandwidth team, we are likely to get
|
|
bottlenecked in one of these areas. I would like to work with the team during
|
|
the summit to help prioritize an item for each dimension so that everyone is
|
|
aware and thinking of what's coming their way for reviews. This will be a
|
|
discussion that involves themed focus to help establish a good review cadence.
|
|
|
|
3. Enablement
|
|
Always! I've always been a fan of enabling folks to do their job, discover
|
|
their interest in the project and help move OpenStack as a whole in a positive
|
|
direction. I think processes help but sometimes they become too hard to deal
|
|
with. I believe the most productive environments are the ones where people who
|
|
are driven & active get opportunities to take important decisions in their
|
|
focus areas while being hit with appropriate trade-offs from partners and the
|
|
vantage point is provided by the PTL.
|
|
|
|
4. Collaboration
|
|
Glance being a service consumed by many OpenStack services, we are a team that
|
|
needs to be actively engaged in cross-project meetings, initiatives, and ad-hoc
|
|
syncs. Import refactor work, deprecation of v1 API and building of
|
|
compatibility layer for Glance requires a ton of cross-project involvement with
|
|
DefCore committee, Nova team, etc. Also, the momentum Glare is getting with its
|
|
consumers will need quite active participation in the respective syncs. I have
|
|
recently taken a co-lead role for the cross project effort on quotas mainly
|
|
with the objective of shaping the right quotas semantics for Glance but also to
|
|
help the bigger OpenStack community in the process. I will give up this co-lead
|
|
role and help the members with a smooth transition and keep myself involved
|
|
from purely Glance perspective to ensure my full attention to the project.
|
|
However, I hope to achieve more collaboration with the help of other team
|
|
members in all the different areas.
|
|
|
|
5. Finding equity to the unbalance
|
|
I believe we've had a good number of drives in OpenStack, developer driven
|
|
cycles, operator driven cycles, and user-driven cycles. There's a ton of value
|
|
to each of these three legs of the tripod on which OpenStack services stand.
|
|
Nonetheless, in a core OpenStack service like Glance, I strongly believe that
|
|
we need to find the center of gravity of this tripod in order to balance the
|
|
development dance. In Newton, I plan to start a broadcast communication channel
|
|
for the use of all the three legs namely developers, operators, and users. We
|
|
already have the MLs and IRC media of communication that fulfill the unicast
|
|
and multicast requirements. I find blogs to be more representatives of personal
|
|
opinions rather than a medium of project updates. I plan to interview the
|
|
active developers, interested operators & users on their work & perspectives;
|
|
publish the update transcripts publicly on non-textual media for a hands-free
|
|
consumption. The most important reason for this proposed initiative being
|
|
communication of intent and not just of the information, and my observation is
|
|
that lack of communication of intent is an important reason for many livelocks
|
|
in the OpenStack people problem realm.
|
|
|
|
[1] https://etherpad.openstack.org/p/newton-glance-summit-planning
|
|
|
|
Thank you for reading and for your consideration.
|
|
|
|
Nikhil Komawar
|