All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
insert inflammatory v6 thread
full disclaimer, thread was inspired by @Maounique
do you guys actually think ipv6 has any chance of being relevant within the next 25 years?
realistically the only actual practical use right now is by mobile providers using nat, which is still linked to ipv4.
i was talking about it in the buyvm discord a while back, so here's a relevant quote:
@Francisco said:
@Maounique said:
@Pixels said: This also explains why we are still at 3.24% IPv6 adoption rate.
I think google's statistics might be misleading because, for example, I disable IPv6 on many devices and mono-purpose machines as well as for family members which are not computer savvy. I have a complex firewall and filtering on ipv4 (mainly for android devices which might call home) and I am not willing to complicate my life replicating that on IPv6 for now.
Googles numbers are misleading because they're heavy mobile. Same with Facebook/twitter.
You want to know the real v6 uptick? Go check any major IX (AMS-IX, NYIIX, Seattle IX, etc) and see if they have protocol breakdowns. Both AMSIX & SIX show V6 at ~7% usage.
There's a very good chance that most of that is just google/CDN traffic anyway.
Actual V6 usage is probably like 1 - 2%.
Francisco
key point:
major internet exchanges are averaging roughly 7% v6 use.
ipv4 keeps appreciating.
you can still get blocks from arin by signing up for the wait list.
i could use only ipv4 and have no issues.
90% of my sites traffic easily comes from ipv4.
tldr:
is ipv6 really even relevant as of today?
state ur two pennies
p.s. sorry @Harambe
Comments
it's of no importance to me
For my website https://yoursunny.com , 9% of HTTP requests were received over IPv6.
The log file covers the period between 2021-09-11 and 2021-10-13.
yeah, but your site is advertised to a very niche audience.
which is surprising since i thought the number would be higher in his case
Not really.
During the same date period, my traffic sources are:
people who follow you know you from the IT crowd, people who are direct linked are obviously from the IT crowd, people who are searching for you are searching for your work in the IT space who would also be in the IT crowd.
I don't get Fran's 'dunk' that v6 is heavy mobile - everything is heavy mobile, most people don't have computers at home and just have an iPhone/Android or maybe a tablet if you're lucky.
I work on very large websites in my day job, if a site is under 60% mobile visits we're checking for issues.
And mobile traffic is heavily v6 at this point because they need the address space. Major carriers are running v6-only stacks because they've run out of RFC1918 space. It addresses their addressing problem and allows them to continue to scale.
It's not the largest %'age of traffic flows, but it's currently a significant chunk of eyeballs and will continue to grow.
I agree with him on his timelines though, v6 uptake is very very slow. It'll easily be another 10-20 years until things are v6-heavy or primary.
There's slow uptake in most of NA/EU because of YUGE v4 allocations. No one's feeling the heat to move yet and all the features that were baked into v6 originally as carrots have been adapted for v4 - so there's very little benefit to move unless you're hard up on address space.
Now, why I push for v6 everywhere is largely 'political'. I want an open internet with very low barriers to entry. v4 pricing going to the moon is great and all, but it basically squashes new competition coming into the space. You're looking at $15k to buy a single /24 at the moment, and I have a feeling we'll see that price at least triple in the next decade.
I want new entrants to the hosting market (especially LE* world), I want new ISPs coming online, etc and not just have the incumbents with gigantic v4 allocations takeover the entire show and then jack up pricing. I want people in developing countries to be able to get to the broader internet when their ISPs can't afford more v4 to add to the CGNAT pools.
So, therefore, I'll support v6 wherever I can and won't buy from providers that don't support it.
I must admit that when IPv6 came out I was disappointed a bit. Looked too much of a radical change from the point of view of simply expanding on "numerology" when I believe IPv4 2.0 would have been a bit more sane.
That being said, there will be many devices which will be on the Internet (which is a nightmare come to think of it, btw) and NAT has some limitations.
When IPv6 came out there were few mobile devices fully connected (TCP/IP I mean), today everyone has at least a phone, some people have 2-3 and with multiple SIM cards, they use VoIP and whatnot.
I think in the next 2 decades the adoption will pick up pace, once we go over 50-60% it will go more or less like with SSL adoption.
harambe pls
I agree it's in theory a good idea, but the mass adoption is far, far, away.
I think in the age in which technology moves so fast, 20 years is, indeed, far-far away. By that time even the virtual devices in the metaverse will have their own TCP/IP stack.
There's many systems/games/etc that are hard coded to v4. I swear I read that Steam's main server search isn't V6 compliant.
https://stats.ams-ix.net/sflow/ipv6.html
I'm all for progress on V6. I've added many features to Stallion over the past couple years to promote V6 even more and have even started giving out free /48's for BGP usage within our network (for anycast, etc).
With that being said, I've had multi day outages on V6 with Cogent and even HE's own BGP tunneling services are on life support. We're finally multi provider native in all DC's so that's a non issue, but it just backs my comment that no one cares about V6.
Go have a multi day outage on V4, hell, even a multi hour outage, and you're dick deep in SLA and endless calls.
V6 fucked up by not being backwards compatible. If it was then everyone would have moved onwards to V6 purely as a marketing move. "We're so advanced, we use the latest technologies", etc.
Francisco
Perception is usually different than reality, especially in groupthink like here.
The reality is IT guys over value v6, in real world practical use v4 is still on top and it isn't even close.
This.
I bet they could design a new, ipv4-backwards-compatible ipv7 and it'd rapidly overtake ipv6, become dominant, and ipv6 would be swiftly forgotten.
Look, nobody here has any illusions about overnight adoption, I think the average timeline people think of is 1-2 decades from now on. That is close to forever in the technology world, so I think nobody can be classified as a "fan" or "hater".
There is a consensus more or less it will be a long time, but the writing is on the wall, prices are rising and rising fast.
If there is anything that can drive change, price is probably it.
It will happen and it will happen faster in very densely populated areas where the technology is booming such as India. China has also a big population but the censorship is more difficult over IPv6. Bummer.
Why do you say this is true? I really don't know, but I thought Chinese censorship was significantly higher in the stack.
In other words, other than expanded address space, ipv6 has no other significant technical advantages over ipv4.
Which is why I believe /25's and /26's get approved.
Upstreams could agree to it, and then allow it on a per customer port basis. Many transit providers send a default route along with their full table so they'd have coverage if they don't want to accept /25's and /26's.
Francisco
I am not sure it was designed to have significant technical advantages. It makes routing less fragmented, though, one point, one allocation, instead of 2-3-n like now.
It is, however the IPv6 action plan lead to little adoption outside of LTE and related.
This makes me think there might be some other reasons than simply "enough" ipv4 space available or NAT everywhere.
Yep, pretty much at this point. There were things like DHCP and IPSec that were baked into the original v6 spec, which got added back to v4 because they were useful features there as well.
Doubt.
Create a new spec, get everyone to agree on it, get software implemented into every operating system, get hardware acceleration working for new routers and switches, get everyone to buy said routers and switches, .... It's 20 years later and v6 is a majority of traffic out of pure necessity, just in time for your v6 replacement to instantly take off.
When your new protocol is not backwards compatible with the old one, there has to be a compelling reason to switch to the new one. The major reason given for 6 was that we were running out of 4 addresses. As you can see however when the price goes up a bit sufficient 4 addresses seem to appear.
To get people to switch to 6, some compelling feature is needed.
So true...
@Francisco
I think you are simply a pioneer in the IPv6, offering Routed /48 Prefix is rare among many VPS providers, but the global bandwidth for IPv6 doesn't seem to me to be particularly large or even clogged, I now have IPv6 on my cell phone and PC, and my ISP can provide me with /56, but I have never seen a larger IPv6 Only site, perhaps even less than the .onion.
In addition, IPv6 does not increase the speed of the Internet, which is the most intuitive way, so people inadvertently overlook it.
Or if an ISP offers IPv4 at 1 Mbps and IPv6 at 100 Mbps, then their IPv6 rate will go up.
Of course, router support is also a problem, even some carriers assign /80, which makes it impossible for Android(only support SLAAC).
@yoursunny：
I think your list should be broken down into whether IPv6 is available and whether the prefix size is greater than or equal to /64.
yeah, but braindead people are in every stack, a few years ago I have negotiated with an ISP for a friend of mine (company) to get a /24 and they allocated 24 IPs on 4 or 5 "subnets". Also, was talking with another ISP some years ago asking for IPv6 and they said I only have 1 IP allocated, not 6.
This does not mean there is something wrong with the system, just with the people.
Some VPS providers, like JustHost, will sell each IPv6 for 1 RUB/month, which is genius.
They took the total # of V6 ip's out there, shoved a dollar sign at the front, and kissed themselves on the foreheads.
Francisco
I think @Francisco nailed it. And I'd add my own major criticism, the insane address size.
Which is much to do with people too btw, because people tend to dislike or even reject anything with (significantly) more than 7 of anything. IP4 is 4 address elements and I'd even go as far as strongly suggesting "IP5a" (64-bit) to be written with 4 elements too rather than 8, like so "abcd.1234.5679.abcd".
People need the feeling they can remember their IP(s).
IPv6 feels impractical, is impractical, and is totally oversized. Something very much IP4-like but with 64-bit addresses is what we need - and I've yet to hear a sensible and realistic reason why we allegedly need more IP addresses than molecules in our solar system.
IPv6 is like telling someone whose 4 cores pre-Ryzen PC feels a bit slow that he absolutely needs IBM's largest mainframe plus a 1 Tb fiber to his house.
Accordingly, the result of 2 decades of "explaining", marketing, trumpeting, herding, pushing IPv6 is ... not even 10% real acceptance and use.
You're confusing "mass adoption" with "mass utilization". The mass adoption isn't a debate, it's supported in hardware, software and deployed globally.
This makes no sense. It's like you're completely missing the point that the players with IPV4 have financial reasons to not replace their money maker. That thing about history and repeating itself...
It's frustrating when people don't understand backward compatibility. It's not backward compatible if you need to rewrite every IPv4 stack to work with something new. Having IPv6 pass through IPV4 networks is what makes IPv6 more backward compatible than your silly IPv7 or jsg's silly idea that accomplishes dick all.
SMH, glad you guys are not involved with these committees.
Of course, people said the same about how we couldn't possibly run out of >4 billion IPv4 addresses and of course we'd never need more than 640k of RAM.
Given that it was designed to replace a system that had run out of space, only an idiot would have done anything but hugely overspec it. At the very least, it doesn't seem likely that we'll have to invent IPv8 any time soon...
Nope, there are many places where it is not used because it is not supported. Also /80 is not actually an adoption, it is at best partial, also /112 /128. That is actually utilization but not adoption.
When the ISP is not supporting it, it is not about utilization, I cannot decide to use it or not when it is not supported by my ISP.
If it is supported in MY hardware, MY network, MY software but the link to the Internet is not supporting it, then it is not adopted/supported because even when I would like to use it I can't because it is not supported. When people want to use it but are stopped in a way or another beyond their control, that is when it is not supported, they are not supported in their attempt to use it, the necessary conditions are not met.