[Idr] Re: [Core] draft-haas-idr-bgp-attribute-escape-04 - WG Adoption call (6/2 to 6/16/2026).

[Jeffrey Haas <[e-mail]>]

Aijun, thanks for the support.

On 6/4/26 21:55, Aijun Wang wrote:
>
> Some suggestions for the current version of this document:
>
> 1.1) Should the document title be changed to "BGP Path Attributes Escape"?
>
I had intentionally kept the name slightly more generic since the 
properties of escape are sometimes subsets of individual path 
attributes.  As an example, non-transitive extended communities where 
the scoping wasn't enforced by a BGP speaker ignorant of the extended 
communities feature.

I agree with you that even such cases are a form of escape for a path 
attribute, but the intention of the current wording to is to encourage 
the readers to "look deeper" than the top level attribute as the 
possible problem.

> 2.2) Should the mitigation mechanism also include the local-asfeature 
> being discussed on the mailing list?
>
I don't yet have a good opinion about how interaction of OAD or iBGP 
local-as masquerading should manifest in the attribute escape document.

OAD is still not adopted and new and in the absence of such procedures 
existing RFC4271/7606 procedures would apply.

iBGP local-as is underspecified and similarly has existing RFC4271/7606 
interactions that cover the functionality.

Part of the motivation to start an adoption call for this document is 
that work on OAD will likely overlap the need to be clearer about escape 
procedures, and their mitigations.

-- Jeff