Network requirements
On most networks nothing has to be changed. Every connection is outbound — from the device and from the PC — and no port is opened inwards on either of them.
This page is for the networks where that is not enough: a firewall that permits only named destinations, an address allowlist, or a plant network where outbound traffic is filtered. If a connection fails on such a network, what follows is the list to hand to whoever manages it.
Everything below applies to both ends — the SMX-RNS device on the machine side and the PC running SX-Client — and all of it is outbound.
What to allow
Section titled “What to allow”Required — this much is enough to connect
Section titled “Required — this much is enough to connect”| Destination | Protocol and port | What it is for |
|---|---|---|
simplinx.net | TCP 443 | Portal, setting up a connection, device sign-in |
tn01-1.simplinx.net — 159.69.215.237 | TCP 443 | Relay (Germany) |
tn02-1.simplinx.net — 93.114.98.167 | TCP 443 | Relay (Turkey) |
tn03.simplinx.net — 5.78.127.70 | TCP 443 | Relay (USA) |
*.pool.ntp.org | UDP 123 | Clock sync — a device with the wrong time cannot sign in, see below |
With only this list permitted, a connection is established and works. There is no port range to open and no inbound rule to write.
The clock is part of connecting. The device verifies the certificate of whatever it connects to, and a certificate cannot be verified against a wrong date. So a device whose clock is far out — one straight out of the box, one that has been stored for a long time, or one whose internal clock battery has died — cannot sign in until it has synchronised its time over UDP 123. A device with a working internal clock connects without it; the first boot on a new network is exactly when it is needed.
Recommended — faster, and usually without a relay at all
Section titled “Recommended — faster, and usually without a relay at all”| Destination | Protocol and port | What it gains |
|---|---|---|
| any address | UDP outbound, any port | A direct connection between the PC and the device: lower latency, and nothing of ours in the path |
| the three relays above | UDP 443, TCP 80, UDP 80 | More ways to reach a relay when a direct path cannot be established |
If your policy allows only one of the two lists, take the required one — it works. The recommended one is what makes it fast.
Devices with more than one uplink
Section titled “Devices with more than one uplink”A device can have more than one way out — the wired WAN, Wi-Fi, a tethered phone, a modem. From firmware 4.0 it watches whichever uplink is carrying its traffic, by making the same connection to simplinx.net it already uses for everything else. After three failed checks it moves the traffic to the next uplink, and moves it back when the first one answers again.
Nothing extra has to be allowed for this — it is the TCP 443 already in the required list above. What it does mean is that an uplink which cannot reach simplinx.net will not be used while another one can. So if a device keeps leaving an uplink you expect it to use, that uplink’s own firewall is the place to look.
A device with a single uplink is never checked and never switches.
Direct connections and relays
Section titled “Direct connections and relays”The PC and the device try to reach each other directly first. A relay is used only when a direct path cannot be established — on a network that blocks outbound UDP, for example.
A relay needs nothing beyond outbound traffic to ports 80 and 443. The session data that travels through a relay uses that same connection; there is no second port to open for it.
Allow all three relays
Section titled “Allow all three relays”Which relay a connection uses is not fixed in advance, and it is not always the nearest one. Allow all three. A firewall that permits only one of them will work until the day it does not — and the failure looks like a device that simply will not connect.
About the addresses
Section titled “About the addresses”Host names can change. The IP addresses change only with an announcement, and when a relay is added or replaced this page is updated with it. Check the date at the bottom before copying the list into a firewall, and prefer the host names where your firewall can resolve them.
Behind a proxy
Section titled “Behind a proxy”SX-Client can connect through an HTTP proxy — enter it in the proxy fields in the client’s settings. That is the route for a PC on a network where direct outbound access is not permitted.
If it still does not connect
Section titled “If it still does not connect”| Symptom | Check |
|---|---|
| The device appears offline in the portal | Outbound TCP 443 to simplinx.net — without it the device cannot sign in at all |
| The device is online, but connecting fails | TCP 443 to all three relay addresses. One of them being blocked is the usual cause |
| It worked before and stopped | A firewall rule listing a single relay by IP. Add the other two |
| It connects, but it is slow | Outbound UDP — without it every session goes through a relay. See the recommended list above |
| The PC is on a restricted network, the device is not | Set the proxy in SX-Client, or allow the addresses above for the PC as well |
| The device is using its backup uplink although the main one is up | The main uplink cannot reach simplinx.net — check TCP 443 on that uplink’s own firewall. An uplink that is up but cannot get out is exactly what this is for |
| A new device, or one stored for a long time, never appears in the portal at all | UDP 123 for clock sync. Until the device has the right time it cannot verify a certificate, so it cannot sign in. The same symptom appears on a device whose internal clock battery has died |
Last updated: 14 November 2026.