RE : Last call for draft-ietf-tewg-interarea-mpls-te-req-01.txt

Hi Tony

Please see inline

Regards

JL

>-----Message d'origine-----
>De : Tony Li [mailto:[email protected]] 
>Envoyé : mercredi 26 mai 2004 22:37
>À : Adrian Farrel
>Cc : LE ROUX Jean-Louis FTRD/DAC/LAN; [email protected]; 
>[email protected]
>Objet : Re: Last call for draft-ietf-tewg-interarea-mpls-te-req-01.txt 
>
>
>> I accept that full leakage of all TE information in a dynamic way
>> would almost certainly
>> have a major impact on IGP scalbility.
>>
>> I do not accept that some carefully considered leaking of aggregated
>> TE information on a
>> sparse timer basis would necessarily have such a dire impact.
>>
>> This is why I want the requirements draft to tell me what impact on
>> the IGP I am allowed
>> to have and leave CCAMP to derive a solution that fits that 
>> requirement. I don't think it
>> is right to tell CCAMP what solutions it may or may not consider.
>>
>
>
>Operator hat on:
>
>I would very much like to see it be a requirement that no information 
>is leaked
>without specific configuration.  Further, I would very much like to see
>implementations give us reasonable granularity (and reasonable 
>abstraction) in
>what we choose to leak.
>
>As an example, I would like to be able to leak detailed topology 
>information from
>my IS-IS L2 database down into an L1 area.  I would NOT want to leak 
>bandwidth
>information.

I don't see the gain, as regards inter-area TE, 
in leaking detailed topology information from L2 into L1 without leaking any bandwidth information...

>I would want to leak an abstraction of another L1 area 
>down into an
>L1 area, but I would definitely want that to be an 
>abstraction, NOT the 
>full
>database.

Of course, else you come back to a single area network !
But IMHO such abstraction, to be useful, i.e. to allow computing an inter-area TE path,
would require to take into account a large set of constraint combinations, 
and this would likely lead to major impact on IGP scalability. 


>
>I am not interested in perfect optimality.  I am interested in 
>what can 
>be
>achieved with a limited amount of overhead.  It is very clear that 
>there cannot be
>any improvements without some impact on IGP scalability.  

Not so sure, there are schemes that allow computing an optimal inter-area path without adding any byte in LSA/LSP...

>If nothing 
>else, every
>byte in an LSA/LSP counts towards router memory consumption.  Thus, we 
>have a
>tradeoff that we need to make and perfect optimality is not a 
>reasonable goal in
>light of limited scalability.  Good returns can be had by leaking 
>information
>that is topologically "close by" and abstracting distant information.  

See above comments on TE abstraction

>If

>necessary, this can be supplemented by path manipulation along the way.
>
>Regards,
>Tony
>
>