Re: WG Last Call: Use of IGP Metric as a second TE Metric

[Francois Le Faucheur <flefauch@cisco.com>]

All,

At 13:39 10/01/2002 -0500, Jim Boyle wrote:

>This email commences a WG last call on forwarding
>
>http://www.ietf.org/internet-drafts/draft-lefaucheur-te-metric-igp-01.txt
>
>to IESG for consideration as a BCP.  This is an individual work item which
>has been discussed and reviewed within TE WG.  It is not a chartered
>deliverable.
>
>The WG last call ends January 25.
>
>Jim


Venkata's comment on draft-lefaucheur-te-metric-igp-01.txt (sent on Dec 5, 
2001) will be incorporated in the final version to be issued at the end of 
the Last Call (to resolve the comments that may arise from this Last Call). 
For memory, here is the prososed resolution for Venkata's comment:


>>   * Consider GMPLS (draft-ietf-ccamp-gmpls-routing-01.txt
>>     section 5) TE Links . That is, for some links there
>>     may not be any IGP adjacency, so there won't be any
>>     IGP metric.
>
>fair point.
>
>>    In that case how are we planning to solve CSPF for
>>     data/voice etc LSPs with a single TE metric. If not
>>     please add sufficient info to the section 2.3 Constraints.
>
>You're right that 2.3 should be updated to reflect this additional constraint.
>We should add in 2.3 a 3rd item along the line of:
>"(iii) It can only be used on links onto which support an IGP adjacency so 
>that an IGP Metric is indeed advertised for the link. For example, this 
>practise can not be used on Forwarding Adjacencies (see [LSP-HIER])."

Francois