Re: [RRG] Moving forward...

[Brian E Carpenter <brian.e.carpenter@gmail.com>]

On 2008-06-20 03:59, Dino Farinacci wrote:
>> There are at least two dimensions in which the differences between v4
>> and v6 may make a difference to what sorts of solutions are effective
>> / deployable / definable / ...
>> 1) The fact that the IPv6 header has lots of bits means that there are
>> solutions that do not involve encapsulation or information loss which
>> can be considered with v6 that do not apply to v4.
>> 2) The fact that v6 is still, in practice, in very early stages means
>> that there is more willingness to change the system to make it worth
>> having.  And that folks are more willing to look at changes.
>>
>> This does not mean that I want to ignore IPv4.  But it does mean that
>> I think the differences may have an impact on the architectural
>> approach we recommend.  And I would hate to see us declare that we
>> will not consider any approach which can not leverage those differences.
> 
> Joel, since we are staying at the conceptual level, I think from an
> architecture point of view we can come up with one architecture for both
> address families. But I agree with you, at the engineering level there
> are different ways to achieve the architecture, where mechanisms can be 
> optimized for the specific address-family.
> 

+1

On 2008-06-20 06:08, William Herrin wrote:
...
> 1. Any valid solution set must support IPv6.
> 
> 2. A valid solution set may but is not required to support IPv4.
> 
> I suspect we have a strong consensus on point 1. The reservations,
> hedging and outright rejection offered strongly suggest that we lack
> consensus (rough or otherwise) on point 2.

The reason, if you're correct about point 2, is probably that
we have no certainty whether in the end-game for IPv4 address space,
the DFZ will deaggregate into a horrendous pre-CIDR-like swamp,
or whether it will remain aggregated due to double NAT (as
advocated by draft-shirasaki-isp-shared-addr). I'd say that if
we want to avoid the horrible future that draft would lead to,
we'd better have an RRG solution for IPv4, i.e. be ready for
large scale deaggregation.

    Brian

--
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