The MPLS WG Archive

[Yakov Rekhter <yakov@juniper.net>]

Nabil,

> I have just had a look at the 03 version of the unnumbered links
> draft. Seems like I missed last call so maybe this is too late but...
> I have a few suggestions in relation to the draft:
> 
> 1. It would be nice if a new ERROR_SPEC c-type is defined for 
>    unnumbered links. This ERROR_SPEC could indicate the node 
>    and interface where an error occured. e.g in the case of 
>    admission control failure. In a number of 
>    current RSVP implementations for numbered networks, the 
>    ip address of the interface where failure occured is inserted
>    into the ERROR_SPEC. It would be nice to have similar information
>    for in the unnumbered case. One would then know the exact
>    link where the failure occured.

ok.

> 2. There is no mention of where the IF_ID RSVP_HOP object is defined
>    A reference to section 9.1.1 of
>    http://www.ietf.org/internet-drafts/draft-ietf-mpls-generalized-signaling-07.
txt
>    would probably be useful. 

Agreed.

> 3. From previous discussion on this list, it appears that the 
>    Interface ID field of the unnumbered ERO subobject is intended
>    to represent the incoming interface of the next hop router.
>    If this is the case, it should be explicitly stated at
>    the beginning of section 6.
> 
>    If the intention is to leave it open such that it is either
>    incoming or outgoing interface then a bit from the reserved
>    field of  the sub-object should indicate which of the two 
>    choices is being used.

The intention is to keep it consistent with numbered interfaces

Yakov.