[Idr] draft-lin-idr-sr-policy-headend-behavior-04 - awaiting version -05

[Susan Hares <[e-mail]>]

Changwang, Jiang, Yisong, Mengxiao, Ran:

During the 5/20/2024 interim, your draft was discussed.  The notes from the interim are below.

You need to create a version -05 before I start an adoption call for this draft.
In version -05, please add the following information:


1.   In that draft, please make sure you indicate the focus of the draft.

My understanding is that the draft is focusing on
"H.Encaps" Encapsulation use case for the Reduce case (H.Encaps.reduced)
And the non-reduced case (H.Encaps.A).


  1.  Provide some feedback on why H.Encaps.A is useful in deployments.

We'll start WG adoption call once I receive the -05 version of this draft with these changes.

Cheerily, Sue Hares


===================
Notes from 5/20 interim
Email reference:
https://mailarchive.ietf.org/arch/msg/idr/XQkAVav3Kg23B5UsOO764_CPvOE/

Reference to 5/20/2024 interim notes:
https://datatracker.ietf.org/meeting/interim-2024-idr-04/materials/minutes-interim-2024-idr-04-202405201400-01

Text in minutes:

3. BGP Extensions of SR Policy for Headend Behavior [15 minutes]
draft-lin-idr-sr-policy-headend-behavior-04 [Changwang Lin]

Ketan: What is the value of defining the headend behavior for SR Policy?
(L3 headend and L2 headend).

Changwang: Should we use one TLV instead of 2?

Ketan: What is the motivation for defining this configuration.
It depends on what type of traffic is being steered?
Is this color traffic? If it is L2 traffic?

Changwang: (shows) the H.Encap.A and H.Encaps.Red. slide?

Sue: Why do we need to include these in the Steering SR Policy?
What benefit does it provide?

Changwang: There is a need for a mechanism to configure the [relationship]
behavior for the headend for the SR Policy. Currently the BSID
Policy does not have the forwarding behavior.

Ketan: The binding SID behavior is an SR behavior.  The "H." behaviors are
headend behaviors. These are different from each other.

Joel: Is this helpful for the SR Policy behavior with reduced policy or not?
There is "H.Encaps" and "H.Encaps.Reduced"?  It is specify whether to
use the reduced mode. Is this one of the values of specifying this in
the SR policy?

Mengxiao Chen: Agree that this can help to control the use of the reduced function.

Ketan: Is it one of the functions or the only function?

Mengxiao Chen: In the previous version, we included the insert and encapsulation.
In this version, we have removed the insertion.  Only the encapsulation remains.

Ketan: Not sure whether it is valuable to indicate the reduced or
non-reduced behavior. Indicate encapsulate or insert may have value, but
insert is not adopted yet.  Only Trouble shooting or some other applications
use the "non-Reduced" function.  I thought the previous vesion makes sense,
but I'm not sure about this version.

Sue: If you go forward with this draft, you make sure you indicate what the
focus of the draft is. This needs to be added before we go forward with adoption.

Sue to Joel: Does this need to go to Spring first.
Joel: The behaviors are already defined in SPRING.

Sue: Changwang after you make the change, we'll do an adoption call to
see if others feel the "Reduced" or "Non-Reduced" is appropriate.