Let's not get into debating the local behavior e.g. Implementation
specifics (since one can argue that the router vendors can provide
comparable optimization for host routes as well as non-host routes).

If I as a vendor were really bothered about host vs. non-host
control-plane & data-plane, then I could take the same standard (e.g.
MAP-E) and implement the preferred part with whatever optimization
possible without benefitting the non-preferred part. It really becomes an
implementation (local behavior) choice.

We don't need more of the same thing. :)

Cheers,
Rajiv

-----Original Message-----
From: Qiong <[email protected]>
Date: Friday, November 9, 2012 11:05 PM
To: Ole Troan <[email protected]>
Cc: Softwires-wg list <[email protected]>
Subject: Re: [Softwires] MAP-E 1:1 for HA

>
>
>
>Hi Ole,
>
>
>On Sat, Nov 10, 2012 at 9:41 AM, Ole Trøan
><[email protected]> wrote:
>
>Yiu,
>
>> I have a question for the HA design concept of MAP-E 1:1. The central
>>theme of MAP-E is to make BR as stateless as possible and use Anycast
>>address to identify the MAP-E BR. However, if we use MAP-E 1:1 mode, the
>>operator must have to pre-provision all the
> subscribe rules to all the BRs sharing the same Anycast address for
>reliable HA. This requires operators to carefully plan out which BRs
>support which subscribers. It is because BR is "per-subscriber stateful"
>in MAP-E 1:1 mode. Compared to the MAP-E design,
> HA in MAP-E only requires the operators to use the same set of rules to
>cover the entire domain. IMHO, this contradicts  the original spirit of
>stateless solution and always puzzles me why MAP-E 1:1 bears the MAP-E
>name. MAP-E and 1:1 MAP-E are two completely
> different solutions and target to different deployment scenarios. I
>would love to hear others to comment in the ML how to resolve this issue.
>
>
>all nodes in a MAP domain must have the same rules.
>in 1:1 mode there is only the CE and the BR in the domain.
>
>having an aggregate route e.g. an IPv4 /24 or a host route a /32 doesn't
>mean you need a different RIB implementation.
>
>
>[Qiong] Actually, that's not the same. If we only need /32, there is no
>need to do LPM implementation anymore, only exact matching is needed and
>this is much eaiser than LPM. LPM is not optimized for /32 routing lookup.
>
>The reason that current routers still implement LPM rather not exact
>matching is because /32 is a truely corner case in RIB. However, if /32
>is not a corner case, but is the major requirement for operators, I
>believe more efficient way should be introduced.
>
>Best wishes
>Qiong 
> 
>
>a host route which is what we have in 1:1 mode is just a corner case.
>what is "completely different" about it?
>
>cheers,
>Ole
>_______________________________________________
>Softwires mailing list
>[email protected]
>https://www.ietf.org/mailman/listinfo/softwires
>
>
>
>
>
>
>-- 
>==============================================
>Qiong Sun
>China Telecom Beijing Research Institude
>
>
>Open source code:
>lightweight 4over6: http://sourceforge.net/projects/laft6/
>PCP-natcoord: 
>http://sourceforge.net/projects/pcpportsetdemo/
><http://sourceforge.net/projects/pcpportsetdemo/>
>===============================================
>
>
>

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

Reply via email to