Hi Linda,

The example you mentioned below seem like a capability discovery and an 
appropriate action taken based on that.  A negotiation is different in my 
opinion, for example IPSEC peers negotiate hashing and encryption algorithm to 
establish a secured connection.

Even before we go into this question, we have to clarify I2NSF control plane. 
In my opinion, I2NSF defines the management plane (controller interfaces) for 
NSF. The control plane and data plane of NSF should be out of scope for I2NSF 
(Look at draft https://datatracker.ietf.org/doc/draft-ietf-i2nsf-framework).

The NSF control plane and data plane, is for each VNF/PNF vendor to define and 
as long as NSF provides I2NSF management interface (NSF-facing interface), it 
should be sufficient.

Thanks & Regards,
Rakesh

From: Linda Dunbar <[email protected]>
Date: Monday, September 12, 2016 at 12:59 PM
To: Rakesh Kumar <[email protected]>
Cc: "[email protected]" <[email protected]>
Subject: Why do we have "negotiation" for the "Control Plane"

Rakesh,

Here is a use case to address your following  question: Cloud (Overlay) Network 
needs the underlay network to enforce Rule A & B &C. Not sure of the capability 
of the underlay network, the Overlay Network can send an inquiry to the 
underlay network asking if it supports Rule A &B&C. The underlay network may 
respond with “Only Rule A”. Then the Overlay network can only send “Rule A” to 
be enforced by the Underlay network.


Control Plane:  In the context of I2NSF, the Control Plane is an
      architectural Component that provides common control functions
      to all I2NSF Components, including some or all of the following:
      authentication, authorization, accounting, auditing, and
      Capability discovery and negotiation.

[Rakesh & Anil] Why do we have “negotiation”? What are we trying to negotiate?


Cheers,
Linda


_______________________________________________
I2nsf mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2nsf

Reply via email to