Re: [Gen-art] Gen-art] Review: draft-ietf-pce-pcep-domain-sequence

[Julien Meuric <[e-mail]>]

Hi Joel, hi Dhruv,

Focusing on the Area ID issue, I'd support adding some text along with 
Dhruv's proposal, but stricter (not just deferring to "implementation's 
feeling"). I.e., Area IDs are AS-specific and mustn't cross AS borders; 
AS-local Area IDs may be used inside an AS (without restricting to the 
origin AS).

See you in Yokohama,

Julien


Oct. 30, 2015 - [e-mail]:
>>          Given that Exclude Route Objects are not interleaved with
>>     include Objects, is there a restriction that Area IDs may only be
>>     excluded from paths within a single AS?
>>
>>
>> [Dhruv]: I guess this would depend on the PCE behavior during inter-AS
>> path computation i.e. PCEmay feel the area id subobjectis irreleventand
>> strips from the XRO before sending the request to another PCE
>> ​ or it might keep it intact. ​
>>
>>
>> This would be in s
>> ​pi​
>> ritof RFC 4874 where
>> ​-
>>
>>
>> ​      ​
>> The number of subobjects to be avoided, specified in the
>>     signaled XRO, may be constant throughout the whole path setup, or the
>>     subobjects to be avoided may be removed from the XRO as they become
>>     irrelevant in the subsequent hops of the path setup.
>>
>> We can always
>> ​use ​
>> EXRS in IRO specify the intentions much more clearly.
>>
>> If you agree, we can work on some text to add.
>
> I still can not see how the Excluded Route Object with an Area ID will
> work.  How will a PCE which receives such a request know what AS it
> applies to?  It works fine if the whole path is within one AS.  But if
> this is a multi-AS request, the AS elements, if present at all, are in
> the IRO.
> The most obvious approach would be to declare that the PCE shall assume
> that all Area ID exclusions apply to the origin AS.