Dave, Colleagues,

Please see inline below.

Ben

On 18 Apr 2011, at 15:20, Mcdysan, David E wrote:
>>> I could paraphrase the contents of draft-medved-alto-svr-apis-00 as "Go
>>> implement draft-gredler-bgp-te-00 and here are some reasons why BGP is
>>> good". Therefore the solution proposed is not really
>>> draft-medved-alto-svr-apis-00 it is draft-gredler-bgp-te-00.
>> 
>> Well, I agree, sort of. draft-medved-* is just a motivation document at
>> this time, eg why we want it, what we think are the requirements.
>> draft-gredler-* defines the new BGP encodings but doesn't say much how an
>> ALTO server would leverage that information. draft-gredler-* also is
>> concerned with MPLS-TE, which is way outside the scope of ALTO.
> 
> I think the following summary is more accurate than the above:
> 
> Note that section 4.1.1 is requirements, and states all of them as
> "should" (I.e., SHOULD) and are not mandatory.
> 
> Section 4.1.2 describes A solution that meets many of the desirable
> requirements. 
> 
>> 
>> A third document is needed (or significant updates to the other two) to
>> say how an ALTO server would consume and react to the BGP TE information.
>> 
>> But right now there doesn't seem to be a forum for this discussion. It is
>> not in scope for ALTO yet (my opinion), and it seems that IDR is not the
>> ideal forum.
>> 
>>> If ALTO generates additional requirements that cannot be met by
>>> existing [routing] protocols ALTO could specify those and take them to
>>> the appropriate Routing Area WG to produce a solution.
>> 
>> That is a possible approach.
>> 
>>> I think draft-medved-alto-svr-apis-00 clearly demonstrates that as it
>>> contains no protocol specification at all and leaves all that to
>>> draft-gredler-bgp-te-00 which can be taken up with the IDR WG.
>> 
>> Sigh. It might also clearly demonstrate that the protocol specification
>> work to make BGP TE work specifically with ALTO servers hasn't started
>> yet. And that draft-medved-alto-svr-apis-00 is written as a motivational
>> (use case / requirements) document not as a standards track protocol
>> document.
>> 
>> I guess we might have to agree to disagree.
> 
> I agree with Rich that the use case/ requirements is separable from the
> protocol specifics.
> 
> Ben, would separating these address what appears to be one of your major
> comments?

Yes. I would prefer ALTO WG to focus on:
- What information an ALTO server needs from the network to do its job (which 
may differ by use case).
- What an ALTO server may want (or be required) to do with that information

> I suggest that we first focus on the requirements and where that work is
> best chartered. 
> 

Did you mean to insert a 'not' to make 'and not where that work is chartered'? 
If so, then I'd say that is a fine suggestion and one I heartily agree with :-)

Ben

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

Reply via email to