[address-policy-wg] 2019-07 New Policy Proposal (Default assignment size for IXPs)

Marco Schmidt wrote on 15/10/2019 13:43:
> This proposal aims to change the default IXP assignment size from a /24
> to a needs-based model, with a /27 as a minimum.

This proposal doesn't seem well-thought-out.

The first, and main, issue is that renumbering IXPs is a terrible 
headache and is something that should be avoided if possible.

The sorts of problems that regularly erupt would include:

- loss of service during renumbering, with traffic being forced over 
backup paths at times that may not suit

- ixp participants being forced to undergo network changes outside their 
normal maintenance window procedures (some networks cannot do this).

- problems with IXP prefix misorigination due to filters not being 
updated properly, which can cause loss of service problems.

- other larger-scale configuration issues due to people making 
substantial config changes to their peering routers, which usually takes 
some time to iron out the bugs.  I.e. this is not as simple as a global 
search + replace.

INEX was a good internet citizen and started out with a /27 on our main 
peering LAN in 1996.  When that ran out, we moved to a /26 and then a 
/25.  We're now at /23.  For each renumbering operation, we ran into the 
problems above, and a lot more.  So from multiple experience, I wouldn't 
wish it on anyone to have to go through an IXP renumbering without good 
reason.  It really is a thorough pain, especially for the IXP participants.

The second issue is that there are ~220 IXPs operating in the countries 
in the ripe ncc region.  This number is pulled from IXPDB 
(https://api.ixpdb.net/v1/provider/list), and cross-referenced against 
the list of RIPE NCC countries here:

https://www.ripe.net/participate/member-support/list-of-members/list-of-country-codes-and-rirs

A /15 has enough space for 512x/24 blocks, which means that this block 
will probably last indefinitely if the minimum assignment size is /24.

By reducing this to /27, all that's going to happen is that IXPs and 
their participants will be dragged through unnecessary renumbering 
procedures; but it is unlikely to make the block last longer.

I'd like to respectfully suggest that the proposal be dropped.

Nick