Re: [RRG] Arguments against Transport, Translation & Six/One Router

[Robin Whittle <rw@firstpr.com.au>]

Hi Tony,

You wrote:

> Is there anyone  other than Christian Vogt who
> thinks it could be practical?
> 
> Yes.

OK - I assume this means you think Six/One Router is practical, or
could be made so - however defined.

How is it practical for the 2nd translation router (in the
destination provider network, like an ETR in map-encap) to handle
this situation?

  Host A in network NA sends two packets to hosts B and C, both of
  which are in network NB and are served by a 2nd translation router
  R2.

  In order to get to R2, the packets are intercepted and translated
  by R1, presumably in NA.

  This translation involves changing the destination addresses of
  the two packets from B and C to R2 and R2 respectively.

  How can R2 distinguish between these packets it receives and
  know which one to send to B and which to send to C?

I am assuming there is no encapsulation, no options header etc.

It seems the only way this could be done is by R1 sending some
supplementary information to R2, which R2 would need to receive
correctly before it could correctly handle these two packets.  I
don't see how sending extra packets like this could be regarded as
robust and efficient enough to be "practical".


Christian Vogt presumably has powerful reasons for needing deep
packet inspection and modification in order to make Six/One Router
work in the "transition period" - which in fact would last forever.
 Can you imagine a way he could do this without deep packet
inspection and modification - or how any purely translation-based
scheme could work without this?

Likewise his need to alter DNS responses.


> I suggest we eliminate "Translation" and "Transport"
> approaches from consideration, since 19 months after
> the RAWS workshop the only potentially practical
> proposals are all Map-Encap.
> 
> Well, that's a matter of opinion, not a matter of fact.
> 
> Now, can we please stop discussing proposals?

We are discussing eliminating areas of the design space.  In order
to do so, I think it is pertinent to discuss what proposals that
space contains at present, as part of discussing what it could ever
contain.

If you or anyone else would like to debate my critique of Six/One
Router:

  http://psg.com/lists/rrg/2008/msg01283.html
  http://psg.com/lists/rrg/2008/msg00533.html

I think this would be a good step towards establishing whether this
translation approach is "practical" (however defined) and whether it
is feasible for anyone to develop another translation based proposal
in the RRG time-frame which is "practical".

I doubt that I can constructively discuss eliminating or keeping
open certain parts of the design space without discussing the
existence or non-existence of proposals in those spaces, and to what
degree they could possibly be widely adopted.  Since you are
co-chair, I interpret this as an instruction not to discuss specific
proposals as part of discussing whether or not to rule out certain
parts of the solution space - so I may not have much more to say on
this.

 - Robin

--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg