Re: 5M prefixes in the core [was: Re: [ppml] [address-policy-wg] Those pesky ULAs again ]
-
To: Thomas Narten narten@localhost
-
From: bmanning@localhost
-
Date: Wed, 30 May 2007 16:01:32 +0000
-
Cc: Leo Bicknell bicknell@localhost, "'ARIN PPML'" ppml@localhost, address-policy-wg@localhost
On Wed, May 30, 2007 at 11:03:52AM -0400, Thomas Narten wrote:
> Leo Bicknell bicknell@localhost writes:
>
> > Today we think of a 5,000,000 prefix Internet as an impossibility.
> > No hardware could ever do that. However, 20 years on I'm not sure
> > a 5 million route Internet will be surprising to anyone.
>
> Who is the "we" you refer to above/
>
> Actually, quite a few people are worried that a 5M prefix Internet is
> a possibility. There are also debates (i.e., no consensus) that when
> that happens, routers will actually be able to cope with the load in
> practice.
>
hum... given that w/ a /32 "boundary" - there exists the
possibility of 2x32 routing table entries... clearly the
/32 boundary is not to preserve routing table slots.
if one is seriously considering a 1-5m entry routing table
then it becomes important to (proxy) aggregate to the /8 or
/9 level to keep within the 1 to 5m entries.
--bill
|