@andiroid

I didn’t ask what kind of “gimmicks” are possible with IPv6 and what IPv6 brings for you, others, the provider or the world, but rather what measurable, concrete, immediately noticeable advantages for me personally IPv6 has now and today My technically very modest household would benefit if I were to tick the v6 box on the PC and the IB today?

I have asked this question over the last 20 years with countless top experts (first in 2002 with Ms. Silvia Hagen, author of several specialist books and founder of the IPv6 Forum), professors from various universities and top ICT experts from leading manufacturers of network equipment (including Cisco and HP) discussed. The answers were always more or less identical and have long been known to me: address scarcity, security, elimination of NAT/multiple NAT, integrated QoS, more efficient routing, advantages with VPN and direct E2E connections, elimination of the DHCP server thanks to SLAAC, easier implementation global transparent addressing concepts, and, and, and.

Everything is good, right, correct and, on the whole, absolutely necessary. long overdue. But it doesn’t even begin to answer my question above.

Show original language (German)

@hed wrote:

I didn’t ask what kind of “gimmicks” are possible with IPv6 and what IPv6 brings for you, others, the provider or the world, but rather what measurable, concrete, immediately noticeable advantages for me personally of IPv6 now and today in my technically very modest household if I were to tick the v6 box on the PC and the IB today?


You would have access to my v6-only website! 🌍🤩👍


Everything is good, right, correct and, on the whole, absolutely necessary. long overdue. But it doesn’t even begin to answer my question above.


I’m afraid we’re getting lost here. I am not the Messiah here who will convert you… 👋

Show original language (German)

@andiroid

There’s nothing to convert, I’m non-denominational when it comes to IPv.

I simply find it extremely interesting (or, to put it bluntly, worrying) that in the last 20 years, despite countless worldwide contacts with experts from the network industry, no one has been able to answer my simple question without evading it.

I asked the question on the last day of a course on this topic. The course instructor, visibly dismayed, came to me after graduation and said that it was a good and legitimate question, but I shouldn’t have asked it publicly in front of everyone else in the course.

Show original language (German)

@hed wrote:

This is also a way to evade the answer and shut down a technical discussion.


Now I’m not sure who is shutting down the discourse:

I didn’t ask what kind of “gimmicks” are possible with IPv6

😳

[…] that I will only switch when IPv4 is EoL.

🙄

For my part, I only change a technology […] when the new technology brings a clear, immediate, measurable, practical advantage for my purposes.

😣

Because I like to prevent possible difficulties and a consistent IPv6-only network concept is not (yet) possible

🤔

I (still) forego IPv6 as much as possible, except for tests and learning purposes

🤨

I just go one step further and deactivate IPv6 completely.

😶

Sure, IPv6 is the present and the future, but IPv4-only operation will be possible for many years to come without any problems for the end user.

No. Definitely no! 😤

From the end customer’s perspective, a problem is only a problem when they perceive it as such, for example because something no longer works. And this is precisely the case that does not occur thanks to the dual stack.

But with v6 disabled. 🤓

The majority of customers probably don’t care at all whether they have v6 via 6rd, natively or no v6 at all.

Incorrect. 👋

Show original language (German)

@andiroid wrote:

You would have access to my v6-only website! 🌍🤩👍



Not only that, but theoretically you could run dozens of servers within the same network, or am I wrong? That would definitely increase the independence of dozens of services (especially among techies).

Of course, only to those who personally care about it. But I prefer to host my passwords myself rather than on a server somewhere in Luxembourg or Canada 😉

Show original language (German)

Kærar kveðjur - herzliche Grüsse
Dominik

That would definitely encourage the independence of dozens of services (especially among techies).

Yup. And allow further decentralized solutions to sprout. But conditionally static v6 addresses, otherwise I’ll still have this useless Dyn-DNS service on my hands.

Of course only to those who personally care about it. But I prefer to host my passwords myself rather than on a server somewhere in Luxembourg or Canada 😉

As long as you trust the tool, it doesn’t really matter where you host the passwords, I think. Notepad++ or the NAS could also blow the texts you write/save to China. Okay, a bit hypothetical, there may be enough honey pots available, especially when it comes to password managers. But that’s also my opinion: it’s one less risk factor if it’s my server where I store the password files. The main thing is that the server has a v6 address.

😁

Show original language (German)

@andiroid

Each of us has our own view of IPv6, mine may be completely off the mark, but please don’t let that sway you. I just ask you to answer the following simple question, the answer to which I am seriously interested, briefly and succinctly, without detours or evasive maneuvers:

What measurable, concrete, immediately identifiable advantages or benefits does it bring to me personally now and today in my very modest, minimalist “KISS household” in terms of network technology if I tick the v6 tick on the PC and the IB today? What changes/improves/makes it easier regarding my daily use of PC, TV,…?

If you can tell me just 1 advantage that I personally notice, I will switch to v6 or switch to mixed operation, otherwise I’ll leave it alone and won’t write another word about v6 in this thread.

Show original language (German)
  • DomiP has responded to this post.

    hed some advantages of IPv6 would be for example:

    DomiP_0-1673731725301.png

    Whether all of these are necessary in a classic KISS household is of course an open question. However, I personally think that the “Security enhancements” or “Simpler header” improvement, for example, can also be good for end customers.

    But that shouldn’t be a direct reason for a change - this evening.

    But as the IETF writes “The Future is Now” - and at some point - at some point - everyone will have to switch to IPv6…

    Show original language (German)

    Kærar kveðjur - herzliche Grüsse
    Dominik

    12 days later

    Tonight my monitoring system reported various alarms. Apparently the reason was that I lost IPv6 connectivity at 3:00 a.m.

    An analysis actually showed no errors. The prefix was still the same and neither IPv6 nor IPv4 addresses had changed. But I no longer received packages routed.

    I only noticed one thing: the firmware 13.20.26/13.20.18/01923 was installed on the IB2. And yes, I’m sure that happened tonight because I checked the firmware yesterday. The download page still says version 13.01.24 (which I had yesterday) as the last version. There isn’t even a beta to download.

    In any case, I then decided to simply restart the router again. After that, my IPv6 connectivity worked again (also with version 13.20.26).

    Something didn’t go right during the update.

    Oh yes, neither IPv4 nor IPv6 prefix has changed.

    Show original language (German)

    @SkyBeam wrote:

    I only noticed one thing: the firmware 13.20.26/13.20.18/01923 was installed on the IB2. And yes, I’m sure that happened tonight because I checked the firmware yesterday.

    Yes, the mass rollout of FW 13.20.26 is currently underway (staggered).

    Show original language (German)

    ….keep on rockin'

    Hi

    Version 13.20.26 has been rolling out since Monday. A few hundred thousand updates go out every night. An update shouldn’t actually have any impact on IP connectivity. But something can always happen. The whole thing probably would have healed itself within 24 hours.

    The thing with the betas somehow got lost. The benefit was probably too small. And the WEB page with the current version is always only updated after the rollout of the latest version is almost complete. This is also a trade-off between the advantages and disadvantages of updating the page too early or too late.

    Show original language (German)

    @5018 wrote:

    The thing about the betas somehow got lost. The benefit was probably too small. And the WEB page with the current version is always only updated after the rollout of the latest version is almost complete. This is also a trade-off between the advantages and disadvantages of updating the page too early or too late.


    The reasoning is understandable, but I think the order is wrong. In my opinion, the website should clearly be updated first - if necessary with a beta. Because nothing is installed automatically and the people who know what they are doing can install the updates themselves earlier and provide feedback (free testers).

    Conversely, there is now no way for me to help someone I know configure this with the current version if their router hasn’t received the update automatically yet and I can’t even test this version anywhere because I don’t have it anywhere can relate.

    The order that is usually used is:

    1. Release beta version - everyone who uses this version probably knows about possible side effects -> tests by experienced users
    2. Release the final version for manual installation - here everyone can test the released version and possibly find problems that were overlooked during testing (real-life scenarios in customer configuration vs. laboratory conditions). Here too, the users who update manually usually know what they are doing -> testing by experienced users
    3. Automated rollout -> Here the user can no longer choose and has to rely on the software working in his configuration.

    Currently, we’re going straight from the lab to the rollout and not even offering experienced users the option to manually obtain the latest version. In addition, there are no release notes on the download page for the version currently being rolled out. So you get a version that doesn’t even have any documentation that you could find anywhere because you only get the documentation after the rollout - possibly at least.

    So maybe in a few weeks after the rollout is completed I will find out what changes were introduced back then and where I might have to react if something had been changed to the firewall, for example.

    It’s even worse. On the download page for IB2, version 13.00.38 is available for download under the link “Download latest version”. And this despite the fact that the release notes for version 13.01.24 are listed directly below. So you don’t even get the correct or expected version under the link.

    Funnily enough, only on the German download page. The English, French and Italian pages link to version 13.01.24. Apparently they don’t want to offer the current version to German users.

    Overall pretty chaotic.

    I don’t know whether IPv6 would have repaired itself automatically after 24 hours, but I don’t really believe it and I don’t know why the router would suddenly start forwarding IPv6 packets again after 24 hours. Presumably this exact use case was not tested in the laboratory (existing IPv6 clients, possibly with prefix delegation and connectivity check after the firmware rollout without further reboot). Maybe someone would have even reported the error if a beta or the final version had been released. If I had the current version, I could recreate the use case and upgrade again after a downgrade to see whether connectivity problems arise again after the renewed upgrade. However, I currently don’t have this option.

    Show original language (German)

    @SkyBeam

    The process is as follows. A FW version is developed within 2-4 months. During this time, a new version is made available to a group of testers (including many super users from this forum) every 3 weeks and tested. This already covers a lot of cases. In addition, of course, the supplier and internally the router team test all possible use cases.

    As development nears the end, the FW will be distributed to up to 3000 employees (automatically), then in the next step to 10000 customers, then 100000 customers, then full rollout. Feedback is constantly collected and monitored to see whether everything is working or whether errors occur. Unfortunately, some things can only be seen in the full rollout because the errors appear in such small quantities that they are not noticeable in smaller quantities.

    By the way, the procedure is very similar for TV.

    I know too. that the use case you mentioned is available to a tester. There were no problems there.

    Back then, I asked my former team to offer the almost finished versions as a beta for everyone. Unfortunately, there wasn’t much interest from testers and the feedback didn’t really provide any new insights. Therefore the benefit is rather very small.

    However, there are sometimes situations in which, for example, there is an error in a current FW that affects a small group of customers and which is solved in the new FW. Then it might make sense to publish a beta, which has already been done.

    Show original language (German)
    a month later

    I’m not sure if this is correct - but I suspect there is a connection:

    Today (possibly I wasn’t at home earlier) I had a lot of problems with my internet connection. Normal websites only loaded after 2-3 minutes, the My Swisscom App could almost only show me the loading circuit because nothing was running anymore, even the local configuration page of the IB3 was very slow.
    An ookla speed test sometimes showed me “no connection”, but sometimes also 40Mbit/s etc.
    But the local network was at the usual >1Gbit/s.

    My Anschluss is very slow at 12/1Mbit/s, but the 5G booster usually delivers 70/30Mbit/s.

    After rebooting the IB3 things got really weird - the booster was displayed as “Connected”, but didn’t have IPV4 but rather IPV6 - and no details about the mobile phone connection. But I couldn’t reach the booster using the IPV6 address.
    In the diagnostic information, instead of “Connection type: VDSL2 - NR5GNSA” there was only “Connection type: VDSL2 - Unknown” - but the GUI continued to show with green ticks that “everything is OK”.

    IPV6 was disabled in the IB3 settings. When I activated it and reloaded the page after saving, it was deactivated again. But I could now access the booster. After rebooting the booster, it got an IPV4 again and the bonding works again (but currently still only about half of the usual performance).

    Ask:

    1. Why can I activate IPV6 with a warning (bonding does not support IPV6), but as soon as I look again in the settings it is deactivated?
    2. In which cases is the booster assigned an IPV4 or IPV6 address? chosen by the IB3 as the connection destination?
    3. Does this have anything to do with this rollout, or should I start a new thread?

    Firmware version 13.20.26/13.20.18

    Show original language (German)

    Hi

    I would turn off the IB3 and the booster and wait 2 minutes. Then start the IB3 and wait 5 minutes, then the booster. Bonding is not a very simple technology and something may have gotten stuck.

    Is the booster connected directly to the IB3 via LAN cable? So no switches in between? If there is a switch in between, turn it off and restart it between restarting the IB3 and the booster.

    Show original language (German)

    @swissbeetle

    The warning message in the GUI explicitly states that applications via IPV6 do not benefit from the booster’s performance. You can still continue.

    Therefore I assume that the IPV6 connections are only routed via copper - if it is completely unsupported then the warning message is false (should not give an option to continue).

    A few weeks ago the IPV6 checkbox was checked by itself, probably because of an update or through the chatbot help troubleshooting = it can definitely be activated.

    @5018

    I would switch off the IB3 and the booster and wait 2 minutes. Then start the IB3 and wait 5 minutes, then the booster. Bonding is not a very simple technology and something may have gotten stuck.

    I did something like this after I was able to connect to the booster via IPV6 via browser. After that, the bonding worked again, the booster also received a local IPV4 address (previously missing for some reason).

    It’s just a bit tedious that this is currently necessary almost every week `🙂

    Is the booster connected directly to the IB3 via LAN cable? So no switches in between? If there is a switch in between, turn it off and restart it between restarting the IB3 and the booster.

    The booster is attached to 3 switches - 2x managed with the correct VLAN (10Gbit backbone in the house, so to speak), 1x unmanaged.
    Once it runs, it runs - until it hangs up again.

    BUT: I have exactly the same problems when I connect the booster directly to the IB3 with a separate cable (I used it like this for several weeks at the beginning). Even then, an almost weekly reboot was necessary.

    - -

    But then it looks more like a different problem, and not because of IPV6 - but I still don’t understand why IPV4 wasn’t assigned and why I can activate IPV6 (the tick is checked), but it is deactivated again when the page is refreshed .

    Since I don’t know much about IPV6 - if IPV6 is deactivated on the IB3, can IPV6 still be used in the local network?
    If yes: ok. If not: strange that I came up with the booster

    Show original language (German)

    The bonding system is currently being completely renewed (somewhat slowly). Then all restrictions will be removed.

    I’m holding off on remotely diagnosing your network. I find it dubious to be able to help you with this.

    Show original language (German)