The questions that come up most, answered properly rather than briefly. Click a question to open it, or the # beside it to copy a direct link.
FlexRadio make an outstanding radio. In an RF sense the 6000 and 8000 series are as good as it gets, and SmartSDR is a genuinely fine piece of software. Nothing here is a criticism of any of that.
Networking, though, is a different discipline. On an ordinary home connection with a public IPv4 address, SmartLink is easy — it works, and it works well. The trouble starts when the connection is not ordinary. Starlink puts you behind carrier-grade NAT, where a great many subscribers share one public address and nothing can be dialled in from the outside. There is no port to forward, because the port is not yours to forward.
And here is the part we rather like: you do not have to take our word for any of it. FlexRadio say the same thing themselves, plainly and helpfully.
“CGNAT and SmartLink” · community.flexradio.com · 8 September 2026
Mike‑VA3MW — Administrator, FlexRadio Employee, Community Manager
“SmartLink is not designed to work with an ISP that is using CGNAT.”
“For SmartLink, it does NOT have to be a fixed IPv4, just a non‑CGNAT IPv4 address.”
On VPN workarounds, he notes that they exist, but that “they are not so simple”. Asked in the replies whether IPv6 would arrive in 4.3, he answered:
“No, there will be no IPv6 in 4.3.”
community.flexradio.com/discussion/8034256/cgnat-and-smartlink
That is a straight answer, and the right one. Flex are telling you exactly what SmartLink was built for, and what it was not. That is not a failing — it is scope. Note the second line especially: the address does not need to be static. It just needs to be a real IPv4 address that is not behind CGNAT.
Which is precisely what Flexitnode gives it. So let Flex carry on doing what Flex does superbly — the RF, the receiver, the panadapter, the radio itself. Let Flexitnode do the one job it is built for: getting that radio working over Starlink. The appliance reaches out to your node, the node has a genuine non‑CGNAT IPv4 address, and your clients connect to that.
Nothing is patched. Nothing is worked around inside SmartSDR, and the radio is left entirely alone. All we do is give the network the shape Flex designed for in the first place — and then SmartLink simply works, the way it does on a home line.
Flexitnode was built to fill one specific gap: you have bought a Flex, you want to run it from a remote site, and the networking is the thing standing in your way.
That site might be a ranch or a farm, somewhere genuinely off‑grid, a hilltop, a corner of a friend’s paddock. It might just as easily be your own house — the radio stays home on the good antennas while you are away for work or travelling. Either way it is one owner, one radio, and a connection that will not cooperate.
What it is not designed for is a complex multi‑user, multi‑radio installation. We can certainly put a Flexitnode into that situation and it will do its job — but if you are running several radios, several sites, or a roster of operators who each need different access, you will get far more out of it with some consulting alongside the appliance than from the appliance on its own. Those builds have requirements no box can guess at.
Flexitnode can do a great deal. We have kept the offering deliberately simple, because that is the offering: buy it, plug it in, get up and running. That is the goal. We would rather do that one thing properly than ship something that does everything indifferently.
Flexitnode is an appliance — a physical box that goes in at your site, not a piece of software you install on something you already own.
It has two physical Ethernet interfaces, and a four‑port relay board fitted with SPDT (single pole, double throw) relays.
The relays are what give you remote power on the radio and remote PTT, with two channels left spare for whatever else needs switching out there. You work them from a simple authenticated web page — not yet another app to install on your phone. There is a little more about that on the home page.
Yes — there is a good deal of spare capability in there.
Flexitnode is an appliance, and it is deliberately sold as one thing that solves one problem properly. But the box can do considerably more than that, and once yours is in and quietly getting on with the job you bought it for, there is nothing stopping you asking more of it.
The easiest way to see what that looks like is to go and look at what VK1HF does with his. At IKARA, the same appliance also handles:
None of that is a different product. It is the same appliance, doing more. Any of it can be done on yours if you want it.
Over time we plan to offer these as simple add‑on modules you can choose to switch on, rather than something that has to be built for you each time. In the meantime, ask — if Ian has already built it once, that is most of the work done.
Then let us talk.
Flexitnode is the appliance — the part you buy, plug in, and are done with. The stations themselves are a bigger thing altogether, and if what you actually want is one of those, built the way ours are, we are open to that.
Every site is its own problem: what power you have, what the sky looks like, what you want to put in the air, and how often anybody can physically get there. So this is a conversation rather than a price list.
Reach out and let us discuss it.
Absolutely.
Flexitnode makes the radio reachable to anyone who has the credentials. Who gets them is entirely up to you, or up to your club: the committee, a training officer, everyone on the membership list, or two people and nobody else. That is a decision about your club rather than about the equipment, and we have deliberately left it where it belongs.
You are not locked in to that decision either. Change the password whenever you like — after a contest, after a working bee, when somebody moves on, or simply because it has been a while. Nothing needs reconfiguring at the site, and nobody has to drive out there to do it.
One thing worth reading next to this: a club with one radio at one site is squarely what Flexitnode was built for. A club running several radios, or wanting different operators to hold different privileges, is the kind of build where a conversation will serve you better than a box on its own.
No. There is nothing up there for you to manage.
The cloud side is part of the service, and Flexitnode looks after it. It is not a server we hand you the keys to and leave you to run.
Your side is short. Put the appliance in at your site, then tell your equipment to connect using SmartLink. That is the whole of it — no extra configuration, nothing to log in to, nothing to keep an eye on.
If you would like to know more about what sits up there, ask and we will happily tell you. But you will never have to touch it.
Let us be upfront about where we are coming from: we are networking people — CCIE and CISSP level networking people. We think ZeroTier is very cool technology indeed, and none of what follows is a dig at it.
But have you tried using it to solve the problem Flexitnode solves? If it were that easy, a great many more people would already have done it. The threads are out there, and they tend to end the same way.
Our honest view is that this is not really the use case ZeroTier was designed for. It is an excellent overlay network. It was never meant to be a way of making a closed appliance behave.
And suppose you do like it, and you do get it working at the house or at the club. Everything is fine right up until somebody wants to connect with their Maestro. So — how do you load ZeroTier onto a Maestro?
Exactly.
That is rather the whole point. Flexitnode lives at the site and in the cloud, not on your devices. Nothing to install on a laptop, nothing on an iPad, and nothing that needs installing on a Maestro — because the Maestro carries on doing exactly what it always does: log in with SmartLink and go. Flexitnode works better here, and that is precisely why it exists.
Essentially, Tailscale and ZeroTier are the same conversation.
There are real differences between them — different protocols, different models, different companies — and in another context we would happily argue about them for an hour. In this context they are very much the same thing.
Tailscale does have a notably more polished end‑user look and feel, and credit where it is due for that. But the problems are the same problems. It is an overlay built for putting software clients onto general‑purpose computers, and it is very good at that. Your Maestro is still not a general‑purpose computer, and you still cannot put a client on it.
So the answer is the same one: Flexitnode asks nothing of your devices.
IPv6 is the dream. Every device with its own address, no NAT anywhere, nothing to forward — the whole problem this product exists to solve simply ceases to be a problem. It is a good dream, and one day it is how all of this will work.
In 2026 it is not. Too many pieces of the path still fail to cooperate: client software, home routers, carrier networks, and the radio itself. FlexRadio have been clear about their part of it — asked directly whether 4.3 would bring IPv6, Mike‑VA3MW answered “No, there will be no IPv6 in 4.3”, adding that “IPv6 is not a simple solution to the problem”.
Rather than assume the rest, we went and tested it. In September 2026, in lab testing against Starlink, inbound connections over IPv6 did not work — even though IPv6 was not subject to CGNAT at all. So the tidy reasoning that “IPv6 has no CGNAT, therefore IPv6 solves this” did not survive contact with the real thing. The address space was not the obstacle. The path still would not carry an inbound session.
Which is rather the case in point: IPv4 is going to be with us for a while yet. Flexitnode is built for the network we actually have in 2026, not the one we would prefer. And if IPv6 does become a clean path end to end, that will be a good day — it makes our job easier, not harder.