Re: [Community-Discuss] inputs on IPv4 Inter-RIR policy proposals (off-topic)
Dear Lee, Jordi, Thank you for the responses. At 02:42 PM 01-07-2019, Lee Howard wrote: Every study I do is independent, so there's that. :) The following is closer to what I have seen: '"independent" is not easy to find, and is not related to a specific region'. For what it is worth, there are other aspects which could be factored in to gain a better view of whether the savings are worthwhile. Regards, S. Moonesamy ___ Community-Discuss mailing list Community-Discuss@afrinic.net https://lists.afrinic.net/mailman/listinfo/community-discuss
Re: [Community-Discuss] inputs on IPv4 Inter-RIR policy proposals (off-topic)
By the way another reference, this one specific to 464XLAT: NAT64/464XLAT Deployment Guidelines in Operator and Enterprise Networks https://datatracker.ietf.org/doc/draft-ietf-v6ops-nat64-deployment/ This document just ended the last-call, so if nothing is identified as really broken (apart from possible editorial nits), the IESG should pass it really soon to the RFC Editor for publication. I expect this becoming an RFC may be in 1-2 months or so. Note that I don't recommend anyone to deploy "only NAT64", the document clearly depicts the whys. Regards, Jordi @jordipalet El 1/7/19 20:11, "JORDI PALET MARTINEZ via Community-Discuss" escribió: Hi SM, Responding from my own perspective, hopefully is useful. I think "independent" is not easy to find, and is not related to a specific region. But here are some "numbers" based on my own experience (from my customers networks, so independent on the sense of "not just one case"). This also follows numbers that other operators (not just my customers) have confirmed, for example in v6ops and other foras. If you deploy CGN (so you want to stay with IPv4 artificially) instead of IPv6, you need CGN boxes to cover 100% of your traffic. Add to this the necessary IPv4 pools for each CGN box. The IPv4 pools will cost money when AFRINIC can't source those anymore. If you deploy an IPv6-only with IPv4-as-a-Service (464XLAT, MAP-T/E, lw4o6, DS-Lite), because the IPv4-only traffic accounts only for the 24%, you only need 1/4 of the "equivalent" box to the CGNs (NAT64, BR, lw4o6, etc.), and probably less than 1/4 of the IPv4 addresses. So, I will bet that you can save in terms of IPv4 addresses and operator infrastructure, with a good knowledge and planning of what are you doing, up to 75%. If your network is mainly residential, because the IPv6 availability in CDNs and caches is in those cases up to 85%, savings can be even bigger. In addition to that, there are some advantages such as the opex savings. It is less expensive to manage IPv6-only with IPv4aaS across your network than pure dual-stack. May be not 50% savings, but still impacting a lot. Some transition mechanisms make a more efficient use of the number of ports (and do not limit how many per customer), so consequently you even need less IPv4 addresses, this is for example the case, in my experience, of 464XLAT. We have some considerations about this at: https://datatracker.ietf.org/doc/draft-lmhp-v6ops-transition-comparison/ (new version coming in a matter of hours ... working on it already!) And by the way, if you need to buy CPEs, please, make sure to ask for RFC8585 support if you want to be safe and save more money! All this may be a bit different depending on the transition mechanism that you choose, and also depending on a broadband-only network, % of residential customers vs business ones (if they require pure dual-stack), cellular or a mix of all those. And of course, what quality of service (number of ports per customer in the CGN or alike) you want to "provide". I know I'm not talking about USD here, because that depends a lot on many factors, case by case: 1) How much you pay for the CGN or alike (NAT64, BR, lw4o6, etc.) boxes. 2) Do you need to replace the CPEs or you have got them already with IPv6 support. 3) May be existing boxes in your network already do some of the CGN or alike (NAT64, BR, lw4o6, etc.) functions. Regards, Jordi @jordipalet El 1/7/19 18:24, "S. Moonesamy" escribió: Dear Lee, At 08:59 AM 28-06-2019, Lee Howard wrote: >A company can save money on IPv4 addresses and CGN by deploying >IPv6. But it's too late to deploy IPv6 before Afrinic runs out of >addresses. Addresses will run out, and the market will not be able >to satisfy the need for addresses. ISPs and mobile carriers >everywhere in Africa will have to deploy CGN, and at higher density >than elsewhere in the world. The cost for businesses to connect will >be much higher, since they need inbound access and therefore unique >IPv4 addresses. African Internet deployment will stall, all because >IPv6 has not been deployed and there is no way to get more IPv4 addresses. Is there any independent study to show that a company in this service region can save money by deploying IPv6? Regards, S. Moonesamy ___ Community-Discuss mailing list Community-Discuss@afrinic.net https://lists.afrinic.net/mailman/listinfo/community-discuss ** IPv4 is over Are you ready for the new Internet ?
Re: [Community-Discuss] inputs on IPv4 Inter-RIR policy proposals (off-topic)
Hi SM, Responding from my own perspective, hopefully is useful. I think "independent" is not easy to find, and is not related to a specific region. But here are some "numbers" based on my own experience (from my customers networks, so independent on the sense of "not just one case"). This also follows numbers that other operators (not just my customers) have confirmed, for example in v6ops and other foras. If you deploy CGN (so you want to stay with IPv4 artificially) instead of IPv6, you need CGN boxes to cover 100% of your traffic. Add to this the necessary IPv4 pools for each CGN box. The IPv4 pools will cost money when AFRINIC can't source those anymore. If you deploy an IPv6-only with IPv4-as-a-Service (464XLAT, MAP-T/E, lw4o6, DS-Lite), because the IPv4-only traffic accounts only for the 24%, you only need 1/4 of the "equivalent" box to the CGNs (NAT64, BR, lw4o6, etc.), and probably less than 1/4 of the IPv4 addresses. So, I will bet that you can save in terms of IPv4 addresses and operator infrastructure, with a good knowledge and planning of what are you doing, up to 75%. If your network is mainly residential, because the IPv6 availability in CDNs and caches is in those cases up to 85%, savings can be even bigger. In addition to that, there are some advantages such as the opex savings. It is less expensive to manage IPv6-only with IPv4aaS across your network than pure dual-stack. May be not 50% savings, but still impacting a lot. Some transition mechanisms make a more efficient use of the number of ports (and do not limit how many per customer), so consequently you even need less IPv4 addresses, this is for example the case, in my experience, of 464XLAT. We have some considerations about this at: https://datatracker.ietf.org/doc/draft-lmhp-v6ops-transition-comparison/ (new version coming in a matter of hours ... working on it already!) And by the way, if you need to buy CPEs, please, make sure to ask for RFC8585 support if you want to be safe and save more money! All this may be a bit different depending on the transition mechanism that you choose, and also depending on a broadband-only network, % of residential customers vs business ones (if they require pure dual-stack), cellular or a mix of all those. And of course, what quality of service (number of ports per customer in the CGN or alike) you want to "provide". I know I'm not talking about USD here, because that depends a lot on many factors, case by case: 1) How much you pay for the CGN or alike (NAT64, BR, lw4o6, etc.) boxes. 2) Do you need to replace the CPEs or you have got them already with IPv6 support. 3) May be existing boxes in your network already do some of the CGN or alike (NAT64, BR, lw4o6, etc.) functions. Regards, Jordi @jordipalet El 1/7/19 18:24, "S. Moonesamy" escribió: Dear Lee, At 08:59 AM 28-06-2019, Lee Howard wrote: >A company can save money on IPv4 addresses and CGN by deploying >IPv6. But it's too late to deploy IPv6 before Afrinic runs out of >addresses. Addresses will run out, and the market will not be able >to satisfy the need for addresses. ISPs and mobile carriers >everywhere in Africa will have to deploy CGN, and at higher density >than elsewhere in the world. The cost for businesses to connect will >be much higher, since they need inbound access and therefore unique >IPv4 addresses. African Internet deployment will stall, all because >IPv6 has not been deployed and there is no way to get more IPv4 addresses. Is there any independent study to show that a company in this service region can save money by deploying IPv6? Regards, S. Moonesamy ___ Community-Discuss mailing list Community-Discuss@afrinic.net https://lists.afrinic.net/mailman/listinfo/community-discuss ** IPv4 is over Are you ready for the new Internet ? http://www.theipv6company.com The IPv6 Company This electronic message contains information which may be privileged or confidential. The information is intended to be for the exclusive use of the individual(s) named above and further non-explicilty authorized disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited and will be considered a criminal offense. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information, even if partially, including attached files, is strictly prohibited, will be considered a criminal offense, so you must reply to the original sender to inform about this communication and delete it. ___ Community-Discuss mailing list Community-Discuss@afrinic.net https://lists.afrinic.net/mailman/listinfo/community-discuss
Re: [Community-Discuss] inputs on IPv4 Inter-RIR policy proposals (off-topic)
Dear Lee, At 08:59 AM 28-06-2019, Lee Howard wrote: A company can save money on IPv4 addresses and CGN by deploying IPv6. But it's too late to deploy IPv6 before Afrinic runs out of addresses. Addresses will run out, and the market will not be able to satisfy the need for addresses. ISPs and mobile carriers everywhere in Africa will have to deploy CGN, and at higher density than elsewhere in the world. The cost for businesses to connect will be much higher, since they need inbound access and therefore unique IPv4 addresses. African Internet deployment will stall, all because IPv6 has not been deployed and there is no way to get more IPv4 addresses. Is there any independent study to show that a company in this service region can save money by deploying IPv6? Regards, S. Moonesamy ___ Community-Discuss mailing list Community-Discuss@afrinic.net https://lists.afrinic.net/mailman/listinfo/community-discuss
