Hi, "IPv4e" -- You probably missed the point about it not needing hardware updates initially at all? Only software. Routing etc. would all remain identical, those extra bits would be in the packet header and only end points need to understand those headers. 52% is not sufficient IPv6 connectivity, even 100% would not be -- It needs to match or exceed IPv4 performance. ie. routing has to be better. Unfortunately, due to design, i don't think that's feasible ever either because the routing tables will grow exponentially larger than IPv4. Right now, IPv6 route is often weaker than IPv4 to my experience. IPv6 OpEx is much higher than IPv4, to our experience as well. It's not even a competition. Br, Aleksi Magna Capax Finland Oy On 01/04/2024 15.37, ivaylo wrote:
The subject of this discussion go a lot ofside. I dont know why you bloat the theme (maybe it is intensionly ?) leading it to IPV4 holding. Our main focus _SHOULD_ be RIPE budget and financing + sustainabla operation in the next year and in the future at all. This of course is related with the members fee.
I done a little research and calculations, to tune my initial propousal so: By IANA public documents current delegated resources to RIPE are:
86016 IPV4 /19 blocks 66624 IPV6 /27 blocks 42882 ASN
If we have a hard coded limits for each LIR, equal steps up, and member fee of 750 EURO per year, each LIR can hold up to:
1 x /19 IPV4 BLOCK (sumary = 32 x /24 networks) 1 x /27 IPV6 BLOCK (sumary = 32 x /32 networks) 16 ASN numbers
After pass one of the above parameters even with one /24 IPV4, /32 IPV6, or ASN, +750 euro (1500 euro RIPE fee) up to the next proporcional limit:
2 x /19 IPV4 BLOCK (sumary = 64 x /24 networks) 2 x /27 IPV6 BLOCK (sumary = 64 x /32 networks) 32 ASN numbers
and so on...
This will generate 64 512 000 euros annual budget for RIPE which is absolutely enough for normal operations in each of the next 10 years, without need to increse members fee each year.
If somebody is afraid, that other RIRs will not keep going with RIPE fees, and there will be posibly leave of the big resource holders, keep in mind that all small members from the other RIRs will move to RIPE. Also good luck moving to ARIN db mess, the risk to lost route of your prefixes is nearly 90%.
---------- Off topic: About IPV4/IPV6 resources, it is imposible to create kind of ipv4 adress space extension. Many hardware do lookup over the ip header, it is not only software related. Even somebody develope a public standart about that, will take decades before all replacing equipment all over the world and another decades all to be workable. We have working IPV6, right now more than 52% has IPV6 connectivity. Biggest EU access operator already give dual stack IPV4/IPV6 even to their end users, and activlly deploy IPV6 stack into their networks. It is much more cheaper in long perspective than to buy/rent IPV4 (NAT/proxy your clients/services over IPV4 and give them IPV6 real addresses). After 2-3 years the core of the Internet will be over IPV6, yes we will need to support IPV4 for the next 10-15 years to provide backward connectivity, but it is a dead end. All who thinks they will do a big profit from IPV4 adresses they hold, do seriously mistake in their logic, but anyway it is their problem. Those who dont implement IPV6 because "yay it is so scary and not working" in one moment will have to do this fast without experience, because will have seroius troubles with IPV4 connectivity.
Ivaylo Josifov VarnaIX / Varteh LTD +359 52 969393 Varna, Bulgaria
_______________________________________________ members-discuss mailing list members-discuss@ripe.net https://lists.ripe.net/mailman/listinfo/members-discuss Unsubscribe: https://lists.ripe.net/mailman/options/members-discuss/aleksi%40magnacapax.f...