On Sun, 10 Feb 2002, Michel Py wrote:
> > What do others think -- is this something worth noting?
> 
> Overall, I find the text excellent.
> 
> I oppose solutions 2,3 and 4 because they bring extra
> complexity to solve a problem that does not exist. The problem
> that does not exist is the need to subnet a /64.

IMO, 2) adds no complexity.  3 might be remotely sensible, but 3 and 4 
both cannot be relied on.. so if you want to be sure, you always have to 
do either 1) or 2).

I hope I made the problems with at least 4) sufficiently clear.

> The reason /127 subnets have become popular is simply because
> of the old IPv4 habit of using a /30 for a point to point
> link. We collectively have to unlearn that.

Well, and then there is also RFC3021, "Using 31-Bit Prefixes on IPv4 
Point-to-Point Links" which is more of the equivalent..

> It is clear that you were reading another thread on this same
> list, "IPv6 Addr/Prefix clarification". There are no reported
> RIR issues, and I still fail to see a valid reason to subnet
> a /64 when one should have used a /48 and subnet using the SLA
> bits. IPv6 is not widely deployed, why can't we just configure
> it right?

At least at one time, Randy Bush advocated /64's for IX's because they 
didn't need more..

If IX's get /48, this is no problem.

Anyway, I see why people see /126 and such as lucrative: they need to
assign only _one_ /64 for all point-to-point links.  There might be e.g. a
couple of hundred of them, and even though 200 of them could fit fine to 
2^16 subnets, you'd still have to define the addressing a bit more 
carefully.
  
> I disagree with the phrasing that solution 1 is a workaround.
> Solution 1 is the way it is supposed to be. Solution 2 is a
> workaround, and not a very good one, IMHO. If an operator has
> to renumber point-to-point links configured with a /126, it
> makes a lot more sense to me to catch the opportunity to
> comply with RFC 2373 and give it a /64 instead.

I didn't see the problem with the wording, but I'll change the text 
slighly to reflect to this.

-- 
Pekka Savola                 "Tell me of difficulties surmounted,
Netcore Oy                   not those you stumble over and fall"
Systems. Networks. Security.  -- Robert Jordan: A Crown of Swords

--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to