This is not a direct reaction to the TAC proposal, but something that's been on my mind in watching the discussion about the TAC proposal unfold on the 44Net mailing list - specifically around what the "goals" should be of... AMPRNet... ARDC... something else?
This community needs to be looking to the future and not the past. The 44Net AMPR "network" is part of the past. It's a key resource, and currently, an exceptionally valuable one based on the current market for large contiguous IPv4 address blocks. Personally, I think selling off part of that IP range was a great move to fund the future. But I think that the future needs to be talked about beyond a reallocating of existing 44Net space and/or how people interact with it (e.g. the apparently now-defunct 44NGN project).
The future of Amateur Radio needs to be in modern technologies and for this community that means IPv6 networks. Specifically, the goal for the future of ARDC/AMPRNet should be that AMPRNet as it exists today is just not that important anymore 5-10 years down the road, and hopefully closer to five. IPv6 has been hyped for decades now and if you've worked in and around networking, IPv6 is "just around the corner" according to the media and industry rags. But what many don't realize is that IPv6 is already here and has been for a long time. For example, Google already sees 35% of its worldwide traffic access using IPv6. The US-based adoption rate is 48%. Japan is 42%. Germany is 52%. India is 61%. My local club sites and personal websites see on average about half of all real connections using IPv6. Mobile phone networks are almost exclusively Carrier Grade NAT for legacy IPv4 + global IPv6 with the IPv6 traffic preferred. I'm sure the other large providers such as Microsoft would see something similar if they published such statistics. Major ISPs here in the US have handed out /64 and /56 IPv6 address blocks to properly configured/enabled equipment by default for at least five years. My point here is that while many pundits keep talking about an IPv6 "cliff event", the adoption of IPv6 was only going to be an over-time uptake and it's been going on in the background for years. IPv6 is very much here, has been for awhile now, and is of paramount importance to emerging countries in regions like APNIC and AFRINIC that don't have massive holdings of legacy IPv4 space like ARIN and RIPE regions do.
ADRC/AMPRNet should be working with and encouraging the amateur radio community to make progress on tools and technologies that render IPv4, and notably the 44.0.0.0/8(ish) allocation, obsolete. Some areas that cry out for an active community leader include:
1. Make useful, educational information available accessible to the amateur operator that helps them understand what IPv6 is, how to use it with their ISP where available, how it's 90% the same as IPv4, benefits of IPv6, etc. I don't think this all needs to be net new material. Perhaps a curation effort with some original content that fills in the gaps important to radio users.
1. Work with hams and the major projects to enable IPv6 natively for common application such as digital voice (D-STAR, YSF, DMR, etc.), Pi-STAR, Allstar Link, IRLP, APRS, etc. This would include providing some funding, maybe in a Google "Summer of Code"-style project, and some expertise. Our area WAN network currently using 44Net space has been dual-stack IPv6 for years. However while our core non-radio systems could all be 100% IPv6, basically nothing about any of the amateur radio stuff we operate even knows what IPv6 is let alone has a prayer of working on it. There's a lot of great stuff out there that was written by someone who wanted to scratch an itch. All that software is great and serves the need, but it requires help and resources to move forward.
1. ARDC/AMPRNet should consider the merits of either outright becoming an LIR[2] and issuing global IPv6 space OR, at minimum, create a formalized "non-collisions registry" of the IPv6 ULA fc00::/7 space[3]. As part of the latter option, it should also define a set of rules for Global/ULA interoperability for "ham purposes" when operators may want their own global IPv6 space to interact with some common/shared ULA space.
1. Work with countries/regions/communities on their radio-based networks. Encourage and support the adoption of IPv6 addressing across these networks including making equipment recommendations, sample configurations, case studies, etc.
1. Help hams understand different protocols that are essential to operation of IPv6 networks. This includes things like SLAAC, DHCPv6, Route Advertisements, Neighbor Discovery, and basic dynamic routing protocols (static routes are wholly impractical in IPv6 networks).
1. ARDC should consider standardized IPv6 capability an essential component of any grant of funds for a project when a project has a networking component.
None of the above is exhaustive, nor any sort of immediately workable plan. But all of this discussion on the re-re-allocation of the 44Net space has had me thinking today about what goals of this community should be and it struck me that we're all talking about a technology from the 1970s that most of the world is actively trying to replace or, at least, deprioritize.
Heading outside now to dig out a bunker....
Jason N8EI
[1] https://www.google.com/intl/en/ipv6/statistics.html [2] https://en.wikipedia.org/wiki/Regional_Internet_registry#Local_Internet_regi... [3] https://en.wikipedia.org/wiki/Unique_local_address
Thank you so much for that e-mail Jason as it’s something that I deeply care about myself!
First, let me add some points to your IPv6 Adoption List:
- The US Government will be IPv6-only by 2030 - The Chinese Government made similar claims, by 2030 - An article on The Register[1] claims that China will go single-stack IPv6 in 2030 and they will bring ~700M users to IPv6 by 2023, plus ~200M IoT devices. There’s another milestone in 2025 (800M people + 400M IoT).
It seems like we’re getting closer and closer to the tipping point now that we see even governments make commitments like that. Especially China that requests the users upgrade their equipment if it’s not capable, and also being a large manufacturer of IoT, could help us get there faster.
Your points are very interesting and similar to what I have shared here before. I would really love to see people here that certainly have the technical knowledge to deploy IPv6, and if there are any issues try to report them, if not fix them upstream. I would also love to see people sharing their knowledge here about what they did, how they did it, what worked, what didn’t, lessons learned, etc.
I would also like to see some funding, similar to what you mention, from ARDC towards these goals. There are many things that could improve and in addition to that, the idea of requiring IPv6 feature equivalence could be interesting to explore.
Finally, as we slightly got into that in another thread today, with my TAC hat on, I can share something:
The TAC is looking at ways to integrate IPv6 into the network and the PoP design so far is fully dual-stack and can support both protocols. This is on top of some other work ARDC is doing, which is also designed with IPv6 in mind. Who knows, maybe part of the ARDC infra (internally) could even be IPv6-only ;)
Antonis
Links: - - [1]: https://www.theregister.com/2021/07/26/china_single_stack_ipv6_notice/
On 28 Jul 2021, at 22:07, Jason McCormick via 44Net 44net@mailman.ampr.org wrote:
This is not a direct reaction to the TAC proposal, but something that's been on my mind in watching the discussion about the TAC proposal unfold on the 44Net mailing list - specifically around what the "goals" should be of... AMPRNet... ARDC... something else?
This community needs to be looking to the future and not the past. The 44Net AMPR "network" is part of the past. It's a key resource, and currently, an exceptionally valuable one based on the current market for large contiguous IPv4 address blocks. Personally, I think selling off part of that IP range was a great move to fund the future. But I think that the future needs to be talked about beyond a reallocating of existing 44Net space and/or how people interact with it (e.g. the apparently now-defunct 44NGN project).
The future of Amateur Radio needs to be in modern technologies and for this community that means IPv6 networks. Specifically, the goal for the future of ARDC/AMPRNet should be that AMPRNet as it exists today is just not that important anymore 5-10 years down the road, and hopefully closer to five. IPv6 has been hyped for decades now and if you've worked in and around networking, IPv6 is "just around the corner" according to the media and industry rags. But what many don't realize is that IPv6 is already here and has been for a long time. For example, Google already sees 35% of its worldwide traffic access using IPv6. The US-based adoption rate is 48%. Japan is 42%. Germany is 52%. India is 61%. My local club sites and personal websites see on average about half of all real connections using IPv6. Mobile phone networks are almost exclusively Carrier Grade NAT for legacy IPv4 + global IPv6 with the IPv6 traffic preferred. I'm sure the other large providers such as Microsoft wo! uld see something similar if they published such statistics. Major ISPs here in the US have handed out /64 and /56 IPv6 address blocks to properly configured/enabled equipment by default for at least five years. My point here is that while many pundits keep talking about an IPv6 "cliff event", the adoption of IPv6 was only going to be an over-time uptake and it's been going on in the background for years. IPv6 is very much here, has been for awhile now, and is of paramount importance to emerging countries in regions like APNIC and AFRINIC that don't have massive holdings of legacy IPv4 space like ARIN and RIPE regions do.
ADRC/AMPRNet should be working with and encouraging the amateur radio community to make progress on tools and technologies that render IPv4, and notably the 44.0.0.0/8(ish) allocation, obsolete. Some areas that cry out for an active community leader include:
Make useful, educational information available accessible to the amateur operator that helps them understand what IPv6 is, how to use it with their ISP where available, how it's 90% the same as IPv4, benefits of IPv6, etc. I don't think this all needs to be net new material. Perhaps a curation effort with some original content that fills in the gaps important to radio users.
Work with hams and the major projects to enable IPv6 natively for common application such as digital voice (D-STAR, YSF, DMR, etc.), Pi-STAR, Allstar Link, IRLP, APRS, etc. This would include providing some funding, maybe in a Google "Summer of Code"-style project, and some expertise. Our area WAN network currently using 44Net space has been dual-stack IPv6 for years. However while our core non-radio systems could all be 100% IPv6, basically nothing about any of the amateur radio stuff we operate even knows what IPv6 is let alone has a prayer of working on it. There's a lot of great stuff out there that was written by someone who wanted to scratch an itch. All that software is great and serves the need, but it requires help and resources to move forward.
ARDC/AMPRNet should consider the merits of either outright becoming an LIR[2] and issuing global IPv6 space OR, at minimum, create a formalized "non-collisions registry" of the IPv6 ULA fc00::/7 space[3]. As part of the latter option, it should also define a set of rules for Global/ULA interoperability for "ham purposes" when operators may want their own global IPv6 space to interact with some common/shared ULA space.
Work with countries/regions/communities on their radio-based networks. Encourage and support the adoption of IPv6 addressing across these networks including making equipment recommendations, sample configurations, case studies, etc.
Help hams understand different protocols that are essential to operation of IPv6 networks. This includes things like SLAAC, DHCPv6, Route Advertisements, Neighbor Discovery, and basic dynamic routing protocols (static routes are wholly impractical in IPv6 networks).
ARDC should consider standardized IPv6 capability an essential component of any grant of funds for a project when a project has a networking component.
None of the above is exhaustive, nor any sort of immediately workable plan. But all of this discussion on the re-re-allocation of the 44Net space has had me thinking today about what goals of this community should be and it struck me that we're all talking about a technology from the 1970s that most of the world is actively trying to replace or, at least, deprioritize.
Heading outside now to dig out a bunker....
Jason N8EI
[1] https://www.google.com/intl/en/ipv6/statistics.html [2] https://en.wikipedia.org/wiki/Regional_Internet_registry#Local_Internet_regi... [3] https://en.wikipedia.org/wiki/Unique_local_address _________________________________________ 44Net mailing list 44Net@mailman.ampr.org https://mailman.ampr.org/mailman/listinfo/44net
On 29/7/21 6:07 am, Jason McCormick via 44Net wrote:
ADRC/AMPRNet should be working with and encouraging the amateur radio community to make progress on tools and technologies that render IPv4, and notably the 44.0.0.0/8(ish) allocation, obsolete. Some areas that cry out for an active community leader include:
- Make useful, educational information available accessible to the amateur operator that helps them understand what IPv6 is, how to use it with their ISP where available, how it's 90% the same as IPv4, benefits of IPv6, etc. I don't think this all needs to be net new material. Perhaps a curation effort with some original content that fills in the gaps important to radio users.
This is related to a suggestion I made earlier - that ARDC could have a role in education, and IPv6 would certainly fit here. I've had native IPv6 myself for over 10 years, and I was able to ensure all of my VPSs have global native IPv6 within a couple of years after that.
- Work with hams and the major projects to enable IPv6 natively for common application such as digital voice (D-STAR, YSF, DMR, etc.), Pi-STAR, Allstar Link, IRLP, APRS, etc. This would include providing some funding, maybe in a Google "Summer of Code"-style project, and some expertise. Our area WAN network currently using 44Net space has been dual-stack IPv6 for years. However while our core non-radio systems could all be 100% IPv6, basically nothing about any of the amateur radio stuff we operate even knows what IPv6 is let alone has a prayer of working on it. There's a lot of great stuff out there that was written by someone who wanted to scratch an itch. All that software is great and serves the need, but it requires help and resources to move forward.
M17 at least supports IPv6 from the outset. My M17 reflector listens on both IPv4 and IPv6, and when I connect using DroidStar, the connection is always IPv6, from what I recall. Reflector based networks like M17 are relatively easy, because if the reflectors run dual stack, the clients can connect using what they have. IPv6 is definitely preferred. Networks with heavy point to point traffic (IRLP, Echolink, AllStar) are much harder, because there's more possibility for IPV4 only and IPV6 only (may have IPv4, but not public) systems to be unable to communicate with each other.
- ARDC/AMPRNet should consider the merits of either outright becoming an LIR[2] and issuing global IPv6 space OR, at minimum, create a formalized "non-collisions registry" of the IPv6 ULA fc00::/7 space[3]. As part of the latter option, it should also define a set of rules for Global/ULA interoperability for "ham purposes" when operators may want their own global IPv6 space to interact with some common/shared ULA space.
There have been proposals for hams to be able to dedicate part of their ISP provided IPv6 space to hame service and have that listed somewhere with ARDC in a registry. For me, that would work, as I have a static /56 (could easily dedicate a couple of /64s to ham use), and they don't filter anything (unless I ask them to - but I've turned the filtering off). But some people may need IPv6 space from elsewhere, to work around ISP restrictions like filtering and dynamic prefixes.
None of the above is exhaustive, nor any sort of immediately workable plan. But all of this discussion on the re-re-allocation of the 44Net space has had me thinking today about what goals of this community should be and it struck me that we're all talking about a technology from the 1970s that most of the world is actively trying to replace or, at least, deprioritize.
I think 44Net will still be important, but encouraging adoption of IPv6 does make sense for looking to the future, as something to do alongside.
I couldnt find the initial proposal but i ask why to separate the two network ? why to make any restriction to any network ? every one will put any rules he want on the firewall he probably have in his network and live with that what is the big need to separate the networks ? i really dont understand Ronen-4Z4ZQ http://www.ronen.org
________________________________ From: 44Net 44net-bounces+ronenp=hotmail.com@mailman.ampr.org on behalf of Tony Langdon via 44Net 44net@mailman.ampr.org Sent: Wednesday, July 28, 2021 8:47 PM To: 44net@mailman.ampr.org 44net@mailman.ampr.org Cc: Tony Langdon vk3jed@vkradio.com Subject: Re: [44net] Future of AMPRNet Should NOT BE AMPRNet
On 29/7/21 6:07 am, Jason McCormick via 44Net wrote:
ADRC/AMPRNet should be working with and encouraging the amateur radio community to make progress on tools and technologies that render IPv4, and notably the 44.0.0.0/8(ish) allocation, obsolete. Some areas that cry out for an active community leader include:
- Make useful, educational information available accessible to the amateur operator that helps them understand what IPv6 is, how to use it with their ISP where available, how it's 90% the same as IPv4, benefits of IPv6, etc. I don't think this all needs to be net new material. Perhaps a curation effort with some original content that fills in the gaps important to radio users.
This is related to a suggestion I made earlier - that ARDC could have a role in education, and IPv6 would certainly fit here. I've had native IPv6 myself for over 10 years, and I was able to ensure all of my VPSs have global native IPv6 within a couple of years after that.
- Work with hams and the major projects to enable IPv6 natively for common application such as digital voice (D-STAR, YSF, DMR, etc.), Pi-STAR, Allstar Link, IRLP, APRS, etc. This would include providing some funding, maybe in a Google "Summer of Code"-style project, and some expertise. Our area WAN network currently using 44Net space has been dual-stack IPv6 for years. However while our core non-radio systems could all be 100% IPv6, basically nothing about any of the amateur radio stuff we operate even knows what IPv6 is let alone has a prayer of working on it. There's a lot of great stuff out there that was written by someone who wanted to scratch an itch. All that software is great and serves the need, but it requires help and resources to move forward.
M17 at least supports IPv6 from the outset. My M17 reflector listens on both IPv4 and IPv6, and when I connect using DroidStar, the connection is always IPv6, from what I recall. Reflector based networks like M17 are relatively easy, because if the reflectors run dual stack, the clients can connect using what they have. IPv6 is definitely preferred. Networks with heavy point to point traffic (IRLP, Echolink, AllStar) are much harder, because there's more possibility for IPV4 only and IPV6 only (may have IPv4, but not public) systems to be unable to communicate with each other.
- ARDC/AMPRNet should consider the merits of either outright becoming an LIR[2] and issuing global IPv6 space OR, at minimum, create a formalized "non-collisions registry" of the IPv6 ULA fc00::/7 space[3]. As part of the latter option, it should also define a set of rules for Global/ULA interoperability for "ham purposes" when operators may want their own global IPv6 space to interact with some common/shared ULA space.
There have been proposals for hams to be able to dedicate part of their ISP provided IPv6 space to hame service and have that listed somewhere with ARDC in a registry. For me, that would work, as I have a static /56 (could easily dedicate a couple of /64s to ham use), and they don't filter anything (unless I ask them to - but I've turned the filtering off). But some people may need IPv6 space from elsewhere, to work around ISP restrictions like filtering and dynamic prefixes.
None of the above is exhaustive, nor any sort of immediately workable plan. But all of this discussion on the re-re-allocation of the 44Net space has had me thinking today about what goals of this community should be and it struck me that we're all talking about a technology from the 1970s that most of the world is actively trying to replace or, at least, deprioritize.
I think 44Net will still be important, but encouraging adoption of IPv6 does make sense for looking to the future, as something to do alongside.
-- 73 de Tony VK3JED/VK3IRL http://vkradio.com
_________________________________________ 44Net mailing list 44Net@mailman.ampr.org https://mailman.ampr.org/mailman/listinfo/44net
On 7/30/21 20:12, R P via 44Net wrote:
I couldnt find the initial proposal but i ask
"You can find our 5-page PDF here: https://pdf.daknob.net/ardc/tac128.pdf https://pdf.daknob.net/ardc/tac128.pdf"
why to separate the two network ? why to make any restriction to any network ? every one will put any rules he want on the firewall he probably have in his network and live with that what is the big need to separate the networks ? i really dont understand Ronen-4Z4ZQ
I still don't either Ronen, I've gone through the emails from today again.
I believe all the use-cases in the proposal could be solved in a much cleaner and less disruptive manner, resulting in a much more modern and flexible solution that could meet everyone's use case, by adopting what PE1CHL proposed two days ago.
There was a suggestion that the current proposal doesn't want to restrict what operators want to do, but it clearly does, it will prevent someone in the 44.128/10 talking to their buddies in the 44.0/9 space.
The obvious answer is that one of them has to renumber, but why should they?
At the moment if 44.155.X.X doesn't want to receive any traffic from non 44 space, some relatively simple iptables rules prevent any traffic reaching them, yet they can still access their buddies on 44 net that do want to receive traffic from.
There was also question/comment about polling and talking to users and co-ordinators. It seems a lot of the co-ordinators on this list were quite obviously not communicated with.
Regards
John
EI7IG
-
Antonios Chariton (daknob) -
Jason McCormick -
John Ronan -
R P -
Tony Langdon