[address-policy-wg] 2017-03 New Policy Proposal (Reducing Initial IPv4 Allocation, aiming to preserve a minimum of IPv4 space)

Hi Riccardo, All,

Thanks for your input.

Please see inline.


On Sun, 24 Sep 2017, Riccardo Gori wrote:

> Dear all,
>
> I started as an ISP early 2015 and I still consider myself a new entrant.

That's not my definition for "new entrant". Strictly i consider a new 
entrant as an organization which doesn't own any IPv4 or IPv6 address 
space. Loosely, it can be a new LIR (before getting its /22 and an 
IPv6 block under current policy)...

But, luckly the community approved a policy for the last /8 long before 
2015. If that didn't happen, your only solution would have been to rely on 
the market.


> In 
> the last 2 years I heard about a couple of time "no more IPv4 policies let's 
> go over and think how to fix/help IPv6 rate adoption" but today we are still 
> here complaining what's the best way to last longer with the agony.

Is "no more IPv4 policies" written in stone somewhere? :-)

Fix IPv6 rate adoption by policy?
Let's call our countries' telecom regulators, and ask for a policy that 
prohibits IPv4 usage from day X onwards? -- i don't think so...........


> For Ipv6 RIPE NCC is doing its best with training and is really appreciated

+1 on that!


> and I learned here that we tend to not mix IPv4/6 policies but I really 
> expected incentives from the cummunity not obstacles. The "IPv6 Requirement 
> for Receiving Space from the Final /8" was abandoned 23/10/2014 by the 
> adoption of 2014-04 proposal

Looking back now, was that a good decision...?


> while this 2017-03 proposal aims to last as 
> longer as possible with IPv4.

Should we tweak it, and make it more "stricter", in a way the address 
space is (automatically?) returned if it's not being used in v4/v6 
translation mechanisms...? (and do we have means to check that?)



> Looks to me that we are trying to save future 
> generation from ice melting saving oil tanks instead of working on research 
> and incentives to clean energies.

Incentives cost money -- taxpayers' or consumers' money (or both!)


> I don't see even any reason to save more address space than the current 
> policies does 'casue we have "trasfert policies"

Transfers of last /8 slices are still allowed after 24 months -- should 
that possibility end...? (2017-03 v1.0 doesn't address transfers)


> for almost any kind of IP 
> resource and if there are some restrictions on new allocation there are more 
> flexible for legacy space.

The RIPE community (through the NCC) only provides services to legacy 
space holders (and there was a proprosal for that... 2012-07).
The RIPE community is not able to design policies regarding address space 
which was distributed before the NCC's creation.



> Today you can simply choose to go RIPE or market 
> as your feeling to get IPv4/6 if needed.

If the choice is going to the "market" (and if you are in strict sense a 
new entrant), the "market" will not advocate you should use/deploy IPv6.
On the other hand, if the choice is becoming a LIR, you will get that... 
:-)



> My small router deals today with more than 2.5 million routes (2 full routing 
> tables and some IX) and it really takes time to backup and even routing 
> performance are affected by volume of routes. I think we should propote IPv6 
> for route aggregation ability.

There is also de-aggregation in IPv6-land. :-(
(http://www.cidr-report.org/v6/as2.0/)
People will mess up routing as the rest of the world lets them...


> I see this policy as:
> - an obstacle to IPv6

That was not the goal. The goal was to extend v4 availability for some 
more time, thus making life easier for IPv6 deployments that will still 
need to talk with the v4 world.


> - a clear side effect of market price rise on IPv4

This proposal is not about the "market". a /24 can cost X, Y or Z over 
time. The only way we can keep "affordability" for new entrants is by 
defining what they get from the NCC, not from any 3rd party.



> - a disincentive to route aggregation

I don't agree. /22s (and bigger chunks) are already being announced as 
/24s. What we consider is that keeping the "affordability" for new 
entrants is a bigger priority than keeping the DFZ on 683k (today) instead 
of 725k, 750k or even 800k routes. I know 800k routes looks insane, but 
two years ago 683k would have been equally insane :-))

ps: On 24-09-2015 (two years ago), 572876 routes
https://web.archive.org/web/20150924225101/https://www.cidr-report.org/as2.0/


Thanks!


Best Regards,
Carlos Friaças


> That's why I oppose this policy
> kind regards
> Riccardo
>
> --
> Riccardo Gori
>