Short version: Sam's statement
> There is a desire to be done with the charter
> sooner rather than later.
seems to be true. The process of revising the
draft LISP WG Charter seems to be so hurried
that no-one has had time to explain why the
LISP list displays 100% consensus that an EID
could be used as an RLOC when this has been
prohibited by draft-farinacci-lisp since its
inception.
Sam wrote:
For the information of the IETF and IESG, I made a consensus
call on the lisp list that there was rough consensus that
there will be cases where the same IP stands both as an EID
and a RLOC.
This is true. Sam's 24 March message was:
http://www.ietf.org/mail-archive/web/lisp/current/msg00336.html
I think there has been enough discussion on-list and other
private comments that the rough consensus of the
participants so far is that there will be cases where the
same IP stands both as an EID and a RLOC.
My claim that this typically does not happen may even be
too strong:-)
I acknowledge that you disagree and that so far our sample
size is small.
Before this message, I only know of one from Noel Chiappa (4
hours before Sam's) asserting that an EID could be used as an
RLOC:
http://www.ietf.org/mail-archive/web/lisp/current/msg00332.html
Noel asserted that it would be technically possible. He did not
state where in the LISP I-Ds or in any other LISP material this
was allowed or required.
Nor did he provide any critique of my argument that it would be
impossible, since an ITR emitting an encapsulated packet with a
destination address XX (in its RLOC role) would probably accept
the just-emitted packet as an EID-addressed traffic packet to
be encapsulated and tunneled to some RLOC address. If the
emitted packet (tunnelled to XX which is also an EID) was not
immediately devoured by the ITR as a traffic packet, then it
would be forwarded to some other ITR, including perhaps one of
the Proxy Tunnel Routers in the DFZ, which would do the same
thing.
I was surprised that this establishment of rough consensus on
the basis AFAIK of one message to the list - and surprised at
the mention of "private comments", apparently as one of the
reasons for this decision.
Robin challenged this consensus call but has not found support
on the list for his challenge.
This is also true.
So, I believe the community disagrees with Robin's claim that
the discussion of LISP identifiers and locators is erroneous.
Judging by the lack of anyone but me disagreeing with Noel's
and Sam's statement this is absolutely true - I agree entirely
with Sam: there has been absolute consensus in agreement with a
statement which I believe violates a central part of the LISP
definition of terms:
http://tools.ietf.org/html/draft-farinacci-lisp-12#page-8
"EIDs MUST NOT be used as LISP RLOCs."
If anyone can explain this, please do.
There is a desire to be done with the charter sooner rather
than later.
Indeed. There was such urgency on the LISP list that no-one has
yet explained how consensus occurred on something which is
apparently so at odds with LISP I-Ds and all the prior
discussion of LISP I know of.
While the draft Charter is being presented for review on the
IETF list, my experience on the LISP list and my perception of
the rejection of other people's suggestions indicates that there
is an extreme reluctance both towards altering the text and
towards detailed discussion of well-meant suggestions for
improving it.
What's the rush?
- Robin
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf