Re: draft-marques-idr-flow-spec-04.txt as an IDR WG document

On Mon, Aug 13, 2007 at 01:21:00PM -0700, Robert Raszuk
wrote:
> Chandra,
> 
> >  While I am personally not against using BGP for a
variety of 
> >applications..
> >  I think the point you raise is very important. It
seems to me that over
> >  time we have used the basic message structure of
BGP and fitted 
> >applications
> >  to them even when all we might have wanted is the
distribution semantics
> >  of BGP. In this case we have constructed an NLRI
that is a long list of
> >  components and atleast _to me_ is a cery creative
interpretation of an
> >  UPDATE message. What about the attributes? There
is a statement that all
> >  standard BGP policy mechanisms should apply as
usual, in the draft but
> >  based on the limited amount of time I have spent
on this draft, I cannot
> >  see that many of the attributes would even apply
here. Till date that has
> >  not been asked for when new AFI/SAFIs are brought
in..
> 
> You are asking the same question again to which I have
already replied 
> on the 10th of Aug. Was there anything in my reply not
clear which would 
> require further explanation ?
> 
> I think I explained well the reasons for flow-spec
rules to be placed in 
> the MP_REACH NLRIs.
> 
> If on the other hand you would like to discuss if new
drafts fit the 
> sematics of BGP UPDATE msg could you start a new thread
and bring some 
> arguments which would justify
draft-pmohapat-idr-info-safi-02.txt to fit 
> semantics of BGP UPDATE msg ? I did spend sufficient
amount of time on 

  I was not specifically pointing to the flowspec idea as
the only one where we are
  crossing a boundary (I do not even know if there is
one)... my comments was 
  that generally that in IDR I have
  not seen that as a community we ask that question when new
AFi/SAFI requests
  are brought in. I agree with you that such a question
might apply to other
  AFI/SAFI also. Paul's point did resonate with my thinking
and my response
  was in the context of his point.

  Rgds,
  Chandra.

> this draft and see no relation of it's proposed NLRI
format to routing 
> information.
> 
> Thx,
> R.

_______________________________________________
Idr mailing list
Idrietf.org
https://ww
w1.ietf.org/mailman/listinfo/idr

Hi Chandra,

And that is exactly why I only suggested you should start a
new thread 
on that topic and let this one be focused on discussion
related to 
flow-spec only.

Cheers,
R.

>  > You are asking the same question again to which I
have already replied
>  > on the 10th of Aug. Was there anything in my
reply not clear which would
>  > require further explanation ?
>  >
>  > I think I explained well the reasons for
flow-spec rules to be placed in
>  > the MP_REACH NLRIs.
>  >
>  > If on the other hand you would like to discuss if
new drafts fit the
>  > sematics of BGP UPDATE msg could you start a new
thread and bring some
>  > arguments which would justify
draft-pmohapat-idr-info-safi-02.txt to fit
>  > semantics of BGP UPDATE msg ? I did spend
sufficient amount of time on
> 
>   I was not specifically pointing to the flowspec idea
as the only one 
> where we are
>   crossing a boundary (I do not even know if there is
one)... my 
> comments was
>   that generally that in IDR I have
>   not seen that as a community we ask that question
when new AFi/SAFI 
> requests
>   are brought in. I agree with you that such a question
might apply to other
>   AFI/SAFI also. Paul's point did resonate with my
thinking and my response
>   was in the context of his point.
> 
>   Rgds,
>   Chandra.
> 
>  > this draft and see no relation of it's proposed
NLRI format to routing
>  > information.
>  >
>  > Thx,
>  > R.
> 

_______________________________________________
Idr mailing list
Idrietf.org
https://ww
w1.ietf.org/mailman/listinfo/idr