Agreed with your statements. When I wrote the previous email, I looked at it
from the 4o6 angle.  I was questioning how was MAP-E more optimized in the
4o6 mode. I over-generalized hub-and-spoke in my question. My bad.

From:  Wojciech Dec <[email protected]>
Date:  Tuesday, March 4, 2014 at 8:41 AM
To:  "Yiu L. LEE" <[email protected]>
Cc:  Softwires-wg <[email protected]>
Subject:  Re: [Softwires] I-D Action: draft-ietf-softwire-lw4over6-06.txt

On 3 March 2014 17:57, Lee, Yiu <[email protected]> wrote:
> How MAP-E aggregates CPE for N CEs in hub-and-spoke? When implementing MAP in
> hub-and-spoke, cpe/ce v4 information is in the br. Each ce will have an entry
> in the br. This is the same number of states lw4o6 will maintain. Am I missing
> something? I support to keep the original text in the draft.

You appear to be confusing the relation between hub&spoke and number of BR
rules.
In MAP optimized + hub&spoke mode the following applies:
- Each CE has just 1 DMR (aka default rule, aka default route). Mesh mode is
not active in this case
- Each BR has just 1 forwarding rule for the N CE's that are part of its
domain. That changes to N rules in the 1:1 case.


Attachment: smime.p7s
Description: S/MIME cryptographic signature

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

Reply via email to