Re: draft-narten-ipv6-3177bis-48boundary-02.txt

["Jason Schiller (schiller@uu.net)" <jason.schiller@mci.com>]

I would like to echo some comments by David Conrad and Mark Smith.  a /48
or 65,536 /64 subnets does seem like an overly large assignment to a
residential user.  

IMHO moving the boundary to a /56 is completely missing the boat on what
the actual problem is.  We discovered that classful addresses were a
mistake and moved to variable length subnet masks.  Let's not repeat this
mistake by not having variable length subnet masks in ipv6.

Now the problem with variable length subnet masks is two fold.  First it
is important that addresses be given out on CIDR boundaries (which is
apparently difficult for some people now that letters are in the
address) so that aggeggation can occur.  Secondly, as Mark points out how
do you strike the right balance between giving a generous amount of space
to allow for growth without the addition of non-contiguous space, while
not being too generous and ignoring conservation?

Can I suggest that the number be based on the expected number of subnets
required in some reasonable time period multiplied by some factor to allow
for growth.  For example, the number of subnets that will be needed in the
next 5 years, rounded up to the nearest CIDR boundary, and the
doubled.  An example of how the math might work is below:

I think I might need 9 subnets in the next 5 years:
1. Trusted home network
2. Wired visitor network
3. Wireless
4. Patio / pool LAN
5. Routing infrastructure
6. Internet facing servers
7. House facing servers
8. Home automation appliances
9. Experimentation

That rounds up to 16 subnets and doubles up to 32 subnets.  That would
require 5 bits for subnetting or a /59

I'm not trying to suggest that this is the right number, but rather that
we should take this sort of approach.

___Jason

==========================================================================
Jason Schiller                                               (703)886.6648
Senior Internet Network Engineer                         fax:(703)886.0512
Public IP Global Network Engineering                       schiller@uu.net
UUNET / Verizon                         jason.schiller@verizonbusiness.com

The good news about having an email address that is twice as long is that
it increases traffic on the Internet.

On Thu, 13 Jul 2006, Mark Smith wrote:
> While I do agree that a /48 per site does seem excessive, particularly
> for a home user, and would therefore accept the /56 per site argument,
> I think it is very important that the balance isn't isn't tipped too far
> back towards conservation when it isn't necessary.

On Thu, 13 Jul 2006, David Conrad wrote:

> Hi,
> 
> Actually, I think Thomas is being very polite here.  I believe the  
> terminology used is more along the lines of "insane", "ridiculous",  
> "ludicrous", etc.  There are also disparaging remarks about the IETF  
> recreating classfull (or even pre-classfull) addressing and repeating  
> history.
> 
> Rgds,
> -drc
> --------
> My opinions are my own and do not necessarily represent the
> opinions of any organization I may be a part of.  So there.
> 
> 
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> ipv6@ietf.org
> Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>