62 lines
2.9 KiB
Plaintext
62 lines
2.9 KiB
Plaintext
My name is Sridhar Ramaswamy and I'd like to announce my candidacy as
|
|
Tacker PTL for the upcoming Newton cycle.
|
|
|
|
Tacker is a newly minted big-tent project. I was involved in Tacker
|
|
right from the days when it used to be a Neutron ServiceVM project and
|
|
during its transition to the NFV Orchestration use case. I was leading
|
|
the project in an informal PTL role over the last two cycles. We had
|
|
the goal of getting into the big tent in Mitaka and few weeks back we
|
|
successfully got accepted - woot! Credit goes to the amazing community
|
|
that believed in Tacker and taken it this far.
|
|
|
|
Mitaka cycle was a quite packed. We are landing many features -
|
|
TOSCA-parser integration, MultiSite VNF placement, Enhanced VNF
|
|
placement (using Nova NFV smarts) and automatic resource creation. We
|
|
also did tons of code clean up to detach Tacker from its original
|
|
Neutron roots. Beyond all this I'm really proud of growing the team
|
|
that is quite diverse for an early stage project [1]. We also
|
|
increased our core team size in Mitaka.
|
|
|
|
Still we have lot more to do in front of us and if elected I'll work
|
|
towards the following goals for Newton:
|
|
|
|
Features: NFV Orchestration is a broad space. We have a lot to build.
|
|
However, we got to pace ourselves as we go through this journey. We
|
|
have many interesting features in the radar like Network Services
|
|
Descriptor (NSD) to orchestrate a sequence of VNFs to stand up a
|
|
network service and VNF Scaling - another highly sought out feature.
|
|
I also support the recent decision to use networking-sfc APIs for
|
|
Tacker VNF Forwarding Graph. I'm confident we would be able to land
|
|
this in Newton.
|
|
|
|
Technical Debt: We did lots of code cleanup over Mitaka cycle. Still
|
|
we mostly focused on getting new features. In the Newton cycle, we
|
|
need to increase our investment in refactoring the aging code that we
|
|
forked off from Neutron. Some of the possible refactorings in our
|
|
sights are in wsgi layer (pecan), in APIs (micro-versioning) and heat
|
|
driver modularity. Decomposing Tacker server into subcomponents is a
|
|
stretch goal which we should get started in the Newton cycle.
|
|
|
|
Release sanity: Now that we are under big tent we need to adhere to
|
|
proper cycle deadlines and milestones. We are learning quickly the
|
|
knobs in the area of release management. I'll work towards getting
|
|
our release pipeline in order in Newton.
|
|
|
|
Gate test: We have a decent amount of gate tests. However, we need to
|
|
keep adding more. I'll work towards increasing the number of gate
|
|
tests particularly using VNFs like OpenWRT.
|
|
|
|
Community: Growing the community is one of the top goals in Newton. We
|
|
need to increase our core team as well. As always I'll do my best to
|
|
welcome new community members, vendors, and operators to join Tacker.
|
|
|
|
Overall we need to balance out our investments between new features
|
|
and technical debt / project stabilization in Newton.
|
|
|
|
Thanks for your consideration.
|
|
- Sridhar
|
|
IRC: sridhar_ram
|
|
|
|
[1] http://stackalytics.com/?release=all&module=tacker-group&metric=patches
|
|
|