[491] in linux-net channel archive
RE: Re: CONFIG_INET_SNARL: What for?
daemon@ATHENA.MIT.EDU (Craig Metz)
Wed Jun 14 10:22:26 1995
To: shields@tembel.org
cc: linux-net@vger.rutgers.edu
Date: Wed, 14 Jun 1995 08:09:36 -0500
From: Craig Metz <cmetz@sundance.itd.nrl.navy.mil>
>> >With CIDR the two high bits are no longer meaningful. There is no longer
>> >any such thing as inherent "class C".
>>
>> Thats not completely true. If your upstream providor aggregates your route
>> then its true, but if you are not part of an aggregate (ie you got your
>> class C more than about 2 years ago) then that logic doesn't hold. There
>> are other reasons that isn't true but I won't go into them.
>It still doesn't work, because in the future blocks will be assigned from
>"class A" space.
Is this postulation, assumption, or known fact?
What is going to happen, IMO, is that classed addresses will be
given out until there are no more, by which time anyone who hasn't moved to
IPv6 will be SOL and will have to mess with address recycling.
>Net 39 is already experimentally split. This means
>that you might get the "class C" 39.1.2, and there is no way to tell from
>the address whether all of 39.x.x.x is at your site or just 39.x.x.(x<16).
But this is somewhat irrelevant, anyway. The whole concept of a
single-homed host knowing what networks are "at your site" or not is FUBAR.
>You cannot tell the size of the block from the address. ifconfig knows
>the correct netmask and that's what should be used.
The interface table holds the correct netmask for all of your direct
attach interfaces. This may be obtained from ifconfig, BOOTP, DHCP, or ICMP
subnet mask advertisements. This has no correlation in a subnetted environment
to your site's classed or CIDR address block.
-Craig
