Re: [mpls] Last Call: draft-ietf-mpls-multicast-encaps (MPLS Multicast Encapsulations) to Proposed Standard

[Eric Rosen <erosen at cisco.com>]

> I believe this document needs a normative reference to 
> draft-ietf-mpls-upstream-label (now in Last Call) because the whole 
> concept of upstream labels is introduced in that document and it 
> seems vital to understanding and implementing this RFC.

Done.

> IANA considerations section should describe how IANA should reword the 
> "unicast codepoint" and "multicast codepoint" assignment fields.  It 
> seems clear that these should be renamed to better reflect the reality 
> introduced in this document.

Done.

> The document does not describe the impact (of lack thereof) of 
> redefining the usage of existing codepoints and modifying existing 
> standards (e.g. the MPLS-in-IP and MPLS-in-GRE encapsulation rules and 
> what may or may not happen in a non-compliant decapsulator).

The new standard is not interoperable with  the old, but we are not aware of
any MPLS multicast deployments that  would be impacted.  I'll add a sentence
or two about this.

> Section 6 gives instructions how to handle unicast and multicast 
> destination address.  It does not describe how to handle the v4 255/8 
> broadcast address and it's debatable which category would apply if a 
> packet was sent to a regular broadcast address such as 192.0.2.255/24. 
> Should this be described for completeness or explicitly declared out 
> of scope?

I would like to declare this to be out of scope.

> Abstract is a bit long and has a typo:
> s/The former "multicast codepoint"/The latter "multicast codepoint"

That's not  a typo.  Instead of  "former multicast codepoint"  I could write
"the codepoint  formerly known as  the multicast codepoint", but  that would
make the abstract longer ;-)

_______________________________________________
mpls mailing list
mpls at ietf.org
https://www.ietf.org/mailman/listinfo/mpls