Re: [tsvwg] draft diffserv-intercon: Handling of a scavenger class / CS1

[Mikael Abrahamsson <[e-mail]>]

On Wed, 17 Jun 2015, [e-mail] wrote:

> As co-chair: I think the short separate standards-track draft would then 
> be the appropriate thing to do. This could be done quickly ***IF*** we 
> get a firm consensus from the community to do this.

If you want to reach out to the "community", I would suggest to for 
instance reach out to the nanog (https://www.nanog.org/list) mailing list 
and ask for feedback there. You'd be more likely to find people who will 
actually implement this there, than on this mailing list. You'd get a much 
wider audience who will tell you if this is deployable in the real world 
with real world equipment and real world configuration, and asking real 
world operators if this is something they could imagine implementing.

Those are the people who will decide what DSCP bit bleaching is done at 
peering borders and customer connections for instance, which is key to 
getting this deployed.

It would also be interesting to hear from equipment vendors, especially 
ones that make BNGs, which kind of bleaching functionality is available in 
existing equipment. Is it possible to just bleach precedence bits on most 
platforms? I'd imagine it is, but it would be interesting to know.

-- 
Mikael Abrahamsson    email: [e-mail]