[mpls] draft-fang-mpls-label-forwarding-no-swap-00

[Robert Raszuk <[e-mail]>]

All,

Few comments, questions on this proposal.

Comment #1:

I very much agree with this draft document that RFC3031 has no explicit
provision for operation which does not require label rewrite. IMO the
update to RFC3031 is needed perhaps by issuing RFC3031bis.

It is implementation optimization how the hardware lookup and header
rewrite are done. But this is not the point here.

It matters to provide right set of label forwarding primitives such that
new protocol work and new architectures which chooses to still use MPLS as
their data plane in non local label scope can be consistently build on top.

As we all recall when RFC3031 was defined some folks were allergic to
anything other then local label significance :)

Therefor I support this document as a base to update RFC3031.


Comment #2:

We already see two individual workarounds and likely we will see more which
try to address the very same problem.

We see H-SDN inventing the term NO_SWAP.

We also see a Segment Routing which cleverly defined a new CONTINUE
operation and simply defined it as MPLS label swap.

"The CONTINUE operation is implemented as an MPLS swap operation."

SRC: draft-ietf-spring-segment-routing-mpls-01


Question 1:

Does this WG wants to see zoo of new terminology across IETF and broader
industry or perhaps ack on the problem and address it ?


Question 2:

What happens if hardware vendor comes with hardware which exposes to
control plane API which allow protocols to signal that mpls label Y is not
to be swapped ?

If so which IETF document such API would be compliant with ?


Best,
Robert.