[38831] in North American Network Operators' Group
Re: standards for giving out blocks of IP addresses
daemon@ATHENA.MIT.EDU (David R Huberman)
Sat Jun 16 19:23:15 2001
Date: Sat, 16 Jun 2001 16:22:32 -0700 (MST)
From: David R Huberman <huberman@gblx.net>
To: Charles Scott <cscott@gaslightmedia.com>
Cc: nanog@merit.edu
In-Reply-To: <Pine.LNX.4.21.0106161709420.19099-100000@harbor.gaslightmedia.com>
Message-ID: <Pine.GSO.4.21.0106161558310.21858-100000@shell1.phx.gblx.net>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Errors-To: owner-nanog-outgoing@merit.edu
Since this is NANOG, I'll restrict my comments to ARIN policy:
> Unfortunately RFC2050 doesn't address at what point additional address
> space should be assigned to an end user, however, it never uses the
> 80% figure in that respect.
Interestingly, ARIN doesn't publish any policy on additional address
space assignments to end-users.
However, I can tell you that in practice (and it's common sense, too) ARIN
does not issue additional assignments to end-users until they demonstrate
that they have used their previous assignment efficiently (80%). Why?
Because:
> This means that if they are already at 50%, you assign them another
> block equial in size to what they have, and they reasonably expect to
> double their utilization in the next year, that everyone should be
> happy.
You can't use 25% of the additional block immediately if you still have
50% of the initial block available*.
Again, though it's not written down (Richard Jimmerson?? Comments?), ARIN
does not issue end-users additional blocks until their existing blocks are
efficiently utilized.
/david
[*] if you can, you're a special case that's not material to this
discussion.
