Post ATTGGCJlFQwMRGTSDo by ceresbzns@infosec.exchange
(DIR) More posts by ceresbzns@infosec.exchange
(DIR) Post #ATTFPvm3VmpO4TmmrQ by carlosefr@mastodon.social
2023-03-09T20:48:58Z
0 likes, 0 repeats
If you follow (even if loosely) the APNIC blog, you've surely came across Geoff Huston's articles. ✍️https://blog.apnic.netI strongly recommend this episode from APNIC's PING podcast where Geoff explains BGP RPKI, and ends up almost making the case *against* it (and IPv6 too).Given he usually covers both topics in a very positive way, I found it awesome how this time he makes you look at these problems from another angle. https://blubrry.com/ping_podcast/90735961/whither-rpki/#apnic #bgp #rpki #ipv6
(DIR) Post #ATTFPxytJ2Hcuro492 by carlosefr@mastodon.social
2023-03-09T21:06:22Z
0 likes, 0 repeats
What I found really interesting, for IPv6 in particular, is the (implicit) notion that we can keep advocating for something over decades using pretty much the same arguments, while the real world changes in ways that make those arguments irrelevant.But that doesn't mean we suddently don't need IPv6. In fact we need it just as much as before, but the most important reason(s) are now different.#ipv6 #advocacy🧵
(DIR) Post #ATTFPyXdDs4Yec3oIK by carlosefr@mastodon.social
2023-03-09T22:00:19Z
0 likes, 0 repeats
We may have to face the fact that end users don't care about IPv6 and maybe never will. They don't care if their devices have a globally routable address. And care even less every day as technically unpalatable stuff like CG-NAT (not giving customers a public IPv4 address at all) sets in without triggering the apocalypse.So, if in a few decades we conclude that killing IPv4 is impossible, we should at least say we're still using it only where it's OK, and IPv6 everywhere else.#ipv6 #advocacy
(DIR) Post #ATTFPzAyrZY2ceTEcy by tschaefer@ipv6.social
2023-03-10T11:12:38Z
0 likes, 0 repeats
@carlosefr sed 's/end users/most end users/g'
(DIR) Post #ATTGGCJlFQwMRGTSDo by ceresbzns@infosec.exchange
2023-03-09T22:04:41Z
0 likes, 0 repeats
@carlosefr As someone who does some self-hosting out of my homelab, it's more of a problem for me that my ISP won't sell me a static IP of *any* kind without upgrading to a business account. Although I suppose a CG-NAT IPv4 would be just as useless as no static IP at all.
(DIR) Post #ATTGGCwkuS8GOCib0C by carlosefr@mastodon.social
2023-03-09T22:11:55Z
0 likes, 0 repeats
@ceresbzns And it's especially sad when ISPs deploy CG-NAT to fixed connections not because they're out of IPv4 addresses, but because they can't manage the ones they have properly (e.g. geographically or stealing them from the mobile side).
(DIR) Post #ATTGGDUmrvM25kdm2y by tschaefer@ipv6.social
2023-03-10T11:22:00Z
0 likes, 0 repeats
@carlosefr @ceresbzns They're out of IPv4 addresses !That's not a question of a better management.
(DIR) Post #ATTH6zX3VzKm8Irmqm by tschaefer@ipv6.social
2023-03-10T11:31:37Z
0 likes, 0 repeats
@carlosefr Geoff Huston is unfortunately lost. He gave up to think that a global internet is a good thing worth it to fighting for.He looks at his statistics - most people only use CDN/cloud based services, so unicast internet is obsolete - his conclusion.He is wrong of course.
(DIR) Post #ATTVeZLQmheI7NKnJo by carlosefr@mastodon.social
2023-03-10T14:14:56Z
0 likes, 0 repeats
@tschaefer @ceresbzns Even though that may be true for most cases, I've heard of CG-NAT deployments happening because the ISPs can't reassign large ranges of unused IPv4 addresses because of the way they structured their network in the past. That's why I said "especially sad".
(DIR) Post #ATTXFKSZBmXIcclS1g by carlosefr@mastodon.social
2023-03-10T14:32:42Z
0 likes, 0 repeats
@tschaefer If that's what the Internet is becoming, shouldn't the fight for the preservation of a global unicast Internet take that into consideration, and shouldn't we update the way we advocate for it?What other arguments can we use in favor of IPv6 that aren't met with the same shrugs that we've gotten from many ISPs over decades?If there's no clear answer, we may be doomed. 😔
(DIR) Post #ATTbUrDG5zdgMeGR5U by carlosefr@mastodon.social
2023-03-10T14:40:52Z
0 likes, 0 repeats
@tschaefer We may not like it, but what used to be temporary workarounds until we have IPv6 everywhere have become competing solutions to IPv6 itself.I'd like to see some numbers on how many mobile ISPs deploy CG-NAT when they run out of IPv4 without deploying IPv6 at the same time. That may be an indicator of the kind of danger we're in.
(DIR) Post #ATTbUrnPvYYwAnBJRo by tschaefer@ipv6.social
2023-03-10T15:20:24Z
0 likes, 0 repeats
@carlosefr most mobile ISPs used CG-NAT from the beginning. At least in France and Germany I know, that all mobile ISPs provide IPv6 in the meantime. JIO in India is pushing up the ipv6 statistics there. Even, one of the worst European countries, Spain, now has startet IPv6 with an significant impact:https://stats.labs.apnic.net/ipv6/AS3352?a=3352&c=ES&x=1&s=0&p=1&w=7
(DIR) Post #ATTgMtGx3PFqCq0nfE by carlosefr@mastodon.social
2023-03-10T16:14:53Z
0 likes, 0 repeats
@tschaefer On the other hand, in Portugal the largest ISP (3243) has had an almost full IPv6 deployment for years on its fixed service.The other largest (2860, roughly as large) has just started ramping up IPv6.Both do CG-NAT on mobile, neither provides IPv6 there (just checked).Vodafone PT also deploys IPv6 to ~60% of its customers but I can't confirm their IPv6/CG-NAT status on mobile.Why is it they seem to be interested in deploying IPv6, but not at all on mobile?