Hi Thinh,
Thanks for the slides. When looking at some of the operations of a VNF (slide 17) it sounds to me that they could be in TOSCA more a workflow than an operation.
Basically, Instantiate VNF somehow would be the install workflow of the substituted node. Same as scale is probably not a node operation but merely a scale workflow that may impact multiple instances within the substitution topology. I guess that most of the
operations of the VNF may end up in being workflows rather than single operation (and somehow a workflow can call a single operation).
I wonder if the substituted node doesn’t expose all it’s workflow to the outside world (seen as operations somehow?) that would be a way of solving the interface
element.
Looking forward for next Tuesday call.
BTW are US switching time this week end?
Luc
From:
"Nguyenphu, Thinh (Nokia - US/Irving)" <thinh.nguyenphu@nokia.com>
Date: Thursday, 9 March 2017 at 19:35
To: Matthew Rutkowski <mrutkows@us.ibm.com>, Luc Boutier <luc.boutier@fastconnect.fr>
Cc: "tosca@lists.oasis-open.org" <tosca@lists.oasis-open.org>
Subject: RE: [tosca] Groups - Issue_TOSCAXXX_Substitution_mappings_grammar_support.docx uploaded
Hi Luc, Matt, and YAML team,
Enclosed attachment is a slide set that was prepared by ETSI NFV on an overview of NFV and VNF Lifecycle management. It very good materials. For our specific topics on deployment flavour,
interfaces, etc, please take a look at slide 10-23. Slide 17 gives an overview of ETSI NFV LCM operations.
I hope this informative materials help the team to understand some nfv requirements.
P.S. the slide set is publicly available on ETSI NFV open area.
Thinh
From: Matt Rutkowski [mailto:mrutkows@us.ibm.com]
Sent: Wednesday, March 08, 2017 9:01 AM
To: BOUTIER, LUC <luc.boutier@fastconnect.fr>
Cc: Nguyenphu, Thinh (Nokia - US/Irving) <thinh.nguyenphu@nokia.com>; tosca@lists.oasis-open.org
Subject: Re: [tosca] Groups - Issue_TOSCAXXX_Substitution_mappings_grammar_support.docx uploaded
Hi Luc,
We just got the use cases described and revisited some of the reasons subst. mappings were created and some of the drawbacks. Primarily we need to understand the use case ("selection" as I call it) and recoginize
that the major problem we acknowledged was that Node Types with custom Interfaces (other operations) when used in subst. mappings have no way of mapping these extra operations (which needs to be considered, especially if they are part of some special state
change). In the case of VNF nodes, they indeed have different names for similar states (and we may ask that they map to ours), but they have additional states as well that we may need to figure out how to acknowledge and communicate to the ORch. where they
are relative to the our standard operational states. Long time ago, we discussed additional grammar to "inject" new states on new interfaces (BEFORE/AFTER a known state).
We will keep discussing next week, but Thinh says it is imperative we have a working solution for the ETSI meetings in 2 weeks (we can have a short term solution or workaround they can document in the NFV profile)
while also seeking a longer term solution for v1.2 or v1.3. We just would like to know what direction any changes might take (and where in the grammar) to better make short term decisions.
Kind regards,
Matt
From: "BOUTIER, LUC" <luc.boutier@fastconnect.fr>
To: Thinh Nguyenphu <thinh.nguyenphu@nokia.com>, Matt Rutkowski/Austin/IBM@IBMUS
Cc: "tosca@lists.oasis-open.org" <tosca@lists.oasis-open.org>
Date: 03/08/2017 08:38 AM
Subject: Re: [tosca] Groups - Issue_TOSCAXXX_Substitution_mappings_grammar_support.docx uploaded
Sent by: <tosca@lists.oasis-open.org>
Hi Thinh, Matt,
I unfortunately was not here on Tuesday and I don’t really understand what this proposal is trying to achieve, it sounds like it is not substitution anymore but a new way to define a node type. How are the
interfaces of the substitution playing vs workflows and how with the nodes within the topology?
I’m not sure to understand what scenario this proposal is solving that classical TOSCA cannot solve. Is there a detailed use-case? Why not having a node template in the topology that defines the interface and
actual capability to be exposed?
Is this going to be discussed again in next Tuesday call ?
Luc
From: <tosca@lists.oasis-open.org> on behalf of Thinh Nguyenphu <thinh.nguyenphu@nokia.com>
Date: Wednesday, 8 March 2017 at 15:18
To: "tosca@lists.oasis-open.org" <tosca@lists.oasis-open.org>
Subject: [tosca] Groups - Issue_TOSCAXXX_Substitution_mappings_grammar_support.docx uploaded
Submitter's message
Hello Simple-YAML team,
This document was presented yesterday YAML call, substitution_mappings grammar enhancement proposal. Thinh
-- Thinh Nguyenphu