[opnfv-tech-discuss] [auto] 11/26 meeting agenda is posted

[Tina Tsou]

Dear all,

Happy Thanksgiving for those who celebrate. Very thankful for all the 
committers, contributors, and whom have already contributed to Auto project, 
making Vexxhost arm64 cloud, intel-pod18, Huawei-pod12 up and running.

11/26 meeting agenda is posted here. It will be held at 6am Monday California 
time as usual.
https://wiki.opnfv.org/pages/viewpage.action?pageId=32014602

We will focus on the following.


AUTO-44
 Paul-Ionut Vaduva

Build ONAP components for arm64 platform



AUTO-71
 Martin Klozik

Implementation of OPNFV scenario for automatic ONAP deployment



AUTO-84
 Paul-Ionut Vaduva

Install OPNFV on Intel-pod18



AUTO-85
 Richard Elias

Install ONAP on Intel-pod18



AUTO-87
 Richard Elias

Re-deploy ONAP at Huawei-pod12


Thank you,
Tina Tsou
Enterprise Architect
Arm
tina.t...@arm.com
+1 (408)931-3833

IMPORTANT NOTICE: The contents of this email and any attachments are 
confidential and may also be privileged. If you are not the intended recipient, 
please notify the sender immediately and do not disclose the contents to any 
other person, use it for any purpose, or store or copy the information in any 
medium. Thank you.
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#22427): 
https://lists.opnfv.org/g/opnfv-tech-discuss/message/22427
Mute This Topic: https://lists.opnfv.org/mt/28291289/21656
Group Owner: opnfv-tech-discuss+ow...@lists.opnfv.org
Unsubscribe: https://lists.opnfv.org/g/opnfv-tech-discuss/unsub  
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [opnfv-tech-discuss] [opnfv-tsc] Discussion of OPNFV Strategic Plan

[HU, BIN]

Manuel,

Thank you for sharing your more thoughts here. I really appreciate your time 
and thinking.

I am not sure if Georg or Ash really meant that we never finished anything 
completely. (Correct me if I am wrong). Here is the status:

  *   We have integrated platform as our main product (through installers). We 
have gone through 7 releases. The stable release process is quite mature. And 
we are working on maturing XCI process.
  *   OVP / Dovetail is also quite stable and becomes mature recently.
  *   The underlying CI/CD pipeline, and related test framework and test cases 
are able to support successful stable release and OVP/Dovetail.

Of course, nothing is perfect, and there is room for further improving 
Integrated Platform and OVP/Dovetail, including add-on features of platform and 
test case coverage in Dovetail. This is more like maintaining and evolving 
existing product. The potential is limited IMHO. The reality is that despite we 
have those 2 flagship products which are very successful and can be maintained 
and evolved further, we are still losing investment significantly as you 
indicated in your 3rd concern.

My take of what Georg and Ash really meant is that we need to be carefully 
handling our existing products (i.e. Integrated Platform and OVP/Dovetail), and 
we don’t lose track of platform capabilities and related test coverage when we 
look for new strategy. The strategy and direction should evolve based on what 
we have and what we are good at. From that viewpoint, I completely agree with 
them. So the strategy of DevOps Platform, which includes the potential 
portfolio that includes existing products, is the evolution based on what we 
have and what we are good at, and opens potential for new market segment:

  *   Expanded market segments:
 *   Existing 2 products are targeted to operators in terms of NFVI 
(current segment)
 *   Potential portfolio expand the segment to all stakeholders and other 
communities in terms of DevOps pipeline (new segment)
 *   Current segment is a vertical of new segment in this picture. So those 
2 segments are orthogonal. New market segment has little cannibalization effect 
on current segment
 *   Thus we expand our addressable market with this new segment, which 
potentially will support unlimited verticals in addition to NFVI vertical.
  *   DevOps Platform, as a horizontal pipeline, is the theme, or a “string”, 
to connect all of our assets and projects together in a systematic way:
 *   Horizontally, DevOps Platform combines our current CI/CD pipeline, 
testing framework, and testing tools with the potential to evolve to a more 
general-purpose pipeline, test framework and tools with the options for 
customization that fits different verticals.
*   We already see the need of evolution to XCI cases
*   DevOps Platforms open the path for further evolution to new 
verticals in addition to NFVI.
 *   Vertically, all feature projects are seen as different verticals of 
the DevOps platform
*   Currently they are all very loose dots
*   DevOps Platform, as a theme and horizontal “string”, connects those 
dots (and new verticals) together in a systematic way
  *   So we have a theme, and a way (i.e. horizontal DevOps Platform) of 
connecting all feature projects, and supporting other verticals in industry

I agree with your suggestion that we need to focus on implementing a couple of 
products at one time. Thus at Step 2 Product Management and Step 3 
Implementation, we need to carefully define:

  *   What are included in portfolio, e.g. those 5 I listed? Or more? Or less?
 *   Being included in portfolio does not mean it must be implemented 
immediately
  *   What is the roadmap and timeline, and how to implement various products 
in portfolio?
 *   How to package each product based on existing assets?
 *   How do we implement new product without impacting existing product 
(i.e. evolution of platform capabilities and test coverage)?
 *   How to evolve/improve existing product to better fit the picture?
 *   What are the gaps?
 *   What are the dependencies and which dependency should be implemented 
first?
 *   etc

The key is the details of how to implement them and when, including what are 
additional platform capabilities and additional test coverage needed to evolve 
in existing products.

Please let me know if you have additional comments and questions. I would be 
happy to discuss more.

Thank you
Bin

From: Manuel Buil 
Sent: Thursday, November 22, 2018 9:33 AM
To: HU, BIN ; Georg Kunz ; Ash Young 

Cc: opnfv-...@lists.opnfv.org; opnfv-tech-discuss@lists.opnfv.org
Subject: Re: [opnfv-tsc] Discussion of OPNFV Strategic Plan

Hey Bin,

Let me try to answer combining a bit both email threads. Thanks for replying by 
the way!

In line with the ROI statement, as you were saying: "we are setting a strategy 
based on potential addressable market and potential 

Re: [opnfv-tech-discuss] [opnfv-tsc] Discussion of OPNFV Strategic Plan

[Manuel Buil]

Hey Bin,

Let me try to answer combining a bit both email threads. Thanks for
replying by the way!

In line with the ROI statement, as you were saying: "we are setting a
strategy based on potential addressable market and potential customer
needs, and our expertise and strength", I am afraid that the different
OPNFV product portfolio which you listed in the previous mail address
different markets and needs. This, in my opinion, will result in not
having a clear focus and thus still not increasing the ROI. I kind of
agree with Georg and Ash that so far we have probably been working on
too many areas (and maybe jumping onto the next new thing) and never
finished any completely. AFAIK, that's what our stakeholders claim,
right? Therefore, I am a bit afraid to jump onto the next new thing
with the devops line of work and that's why I was asking whether you
have some indications pointing to that market as an addressable one for
OPNFV (we definitely have skills in that area!).

My suggestion would be to focus on one thing which could result in 2 or
3 "products". After a few years, I wonder if our "original product" of
OPNFV (the reference patform) is still interesting for anyone
(specially stakeholders) and thus we should focus there. According to
Georg and Ash's mail, it seems they have some indications that it is
and we are not too late. Unfortunately, I can't really tell but I'll
try to investigate a bit :).

Regards,
Manuel

On Wed, 2018-11-21 at 21:34 +, HU, BIN wrote:
> Hi Georg and Ash,
>  
> Thank you very much for sharing your view and concern.
>  
> First of all, please refer to the message I sent earlier that tried
> to answer Manuel’s concerns. That message outlines the thought-
> process and my view to address Manuel’s 3 concerns.
>  
> Then back to your concern, i.e. integrated platform capability and
> compliance toolset that we are having now. I am really sorry that if
> there is a misunderstanding here. My view is that the offering (or
> the delivered product) will be defined in Step 2 in detail (as a
> Product Management function). My personal understanding is that what
> we can offer is a portfolio instead of a single product. Our OPNFV
> product portfolio may include:
> DevOps solution as outlined in User Story on slide #12.
> A packaged testing tool chain that can be offered standalone
> An Integrated Platform Capabilities grown from our DevOps pipeline
> (our original product)
> A conformance testing solution of the integrated platform (our CVP /
> Dovetail)
> A LaaS infrastructure solution
> Etc.
>  
> Just like a cloud provider has a portfolio of products and services,
> including fundamental IaaS, PaaS and SaaS solutions and services. So
> it is a portfolio.
>  
> Please note that I used the word “Product” to  illustrate what we can
> offer. It does not mean anything commercial.
>  
> Hopefully I clarified it. Of course, we cannot do it all at one time.
> so when we define product portfolio in Step 2, we also need to define
> the roadmap and timeline in long term view. Once we have the
> strategy, and then portfolio and roadmap, we will have a very good
> story to market it and attract investment and resources back, and of
> course the most important is to implement it with necessary
> resources.
>  
> Please let me know if you have more questions.
>  
> Thank you
> Bin
>  
> From: Georg Kunz  
> Sent: Wednesday, November 21, 2018 9:04 AM
> To: Ash Young ; HU, BIN 
> Cc: opnfv-...@lists.opnfv.org; opnfv-tech-discuss@lists.opnfv.org
> Subject: RE: [opnfv-tsc] Discussion of OPNFV Strategic Plan
>  
> Hi Bin, Ash, all,
>  
> I’d like to pick up Manuel’s question about the value that our
> stakeholders would like to see and Ash’s point of building a
> reference platform: I believe that both aspects point towards the
> need for strengthening our compliance program – which is already
> based on a broad base of great test tools. OVP does not yet consume
> enough of the available tests for various reasons – having to admit
> this as a former Dovetail PTL. But I also think that we can still
> improve the capabilities of the OPNFV platform – by means of
> integration and closing gaps upstream. This is a requirement for
> adding additional tests covering NFV capabilities to the
> corresponding test tools and then eventually to OVP. Additional NFV
> capabilities we could think about include, for instance, L2GW, SR-
> IOV, LBaaS, FWaaS – in addition to emerging use cases like edge
> computing and cloud native computing, i.e., covering both OpenStack
> and K8s-based deployments.
>  
> So, in the context of the proposed DevOps approach, I am a little
> concerned that we lose track of enabling platform capabilities which
> are a requirement for the test tools and the compliance program. We
> need to make sure that this does not get out of focus too much (in my
> opinion). Specifically, if the main deliverable of OPNFV is an
> integration and CI framework, who do we consider performs the
>