[ppml] Policy Proposal 2002-3

Owen DeLong owen at delong.com
Mon Oct 6 21:47:15 EDT 2003

--On Monday, October 6, 2003 18:15 -0600 Trevor Paquette 
<Trevor.Paquette at terago.ca> wrote:

>> -----Original Message-----
>> From: Calvin Browne
>> Sent: Friday, October 03, 2003 6:06 AM
>> Subject: Re: [ppml] Policy Proposal 2002-3
> ....
>> Why should 2002-3 not be passed? or what are the arguments for not
>> passing it?
> Folks used to be able to receive direct assignments from ARIN (and other
> registries) in the past. This ability was revoked for some reason, and it
> seems that no-one can remember why...
It was not revoked.  I recently got several direct assignments from ARIN.
The convenient boundary for getting an assignment is currently /20.  The
original intent of RFC2050 appears to have been /20, except in some
circumstances which allow a /24.  The /24 ability seems to have been mostly
deprecated by ARIN policy.  2002-3 without modification attempts to move the
current /20 boundary for end-user assignments to /22.

> I'd hate for us to go back down the same path and encounter the same
> problems as did our predecessors only to later go 'DOH!.. no wonder they
> revoked this ability'.
It was never revoked.  It was raised up to /18 and later relaxed to /20.
This was done because of routing table growth and backbone routers that
could not handle the scaling of the routing table.  At the time, routers
were having trouble handling more than 50-70k prefixes.  Today, routers
are handling the 120+k prefixes in today's routing table without difficulty.
There is little to indicate that relaxing this to a /22 at this point would
outstrip the gains in router technology since that time.

> Why was this ability revoked and the current policy put in place?? Can
> anyone explain that?
It was not revoked.  Current policy was put in place as a relaxation from
the initial over-reaction to the situation.  This is a further 
of that towards the original policy now that the technology can handle it.


More information about the ARIN-PPML mailing list