📘 Reflector Setup Guide

A reflector is the server where DVPX users meet. Apps connect to it over TCP, announce their identity and pick a talkgroup; while someone talks, voice packets travel to the reflector over UDP and it forwards them to everyone on that talkgroup.

Reflectors also link to each other: a talkgroup is shared across the whole network, so which reflector you are on never splits a conversation. A reflector NEVER connects to the database — it talks to the dashboard only through the API token issued to it. That is why no database credentials ever live on your server.

1. Sign up. Your callsign becomes your username. Your account waits for administrator approval.
2. File a reflector request. Give your server address, ports and justification. You do not have to rent the server before this step, but you do need to know its address.
   The form also asks for SSH details, but they are OPTIONAL — you may leave them empty. If you provide them, an administrator can look at your server during an outage and help with the setup; the password is stored encrypted and only an administrator can read it.
3. Approval and token. The administrator approves the request and creates the reflector record. The moment they generate a token it waits for you HERE ON THIS PANEL: press ⚿ Show Token in the My Reflectors list on "My Panel" and it appears together with a ready-made config.json. The token does NOT need to be sent over WhatsApp or e-mail — and it should not be.
   While waiting on the panel the token is stored encrypted, with the key kept outside the database. Once installed, clear it with "Got it, clear from my panel".
4. Prepare the server. Install Node.js, place the files, write config.json, open the ports, start the service. The sections below cover this.
5. Verify. Once your reflector starts reporting to the dashboard it shows as LIVE in the "My Reflectors" list on your panel. If it does not, the same row tells you why.

A talkgroup number BEGINS WITH THE COUNTRY CODE. The country you pick when signing up determines that code, and you may only request numbers starting with your own. This splits the number space by country so operators from two countries never ask for the same number.

1) Node.js

Check first: if the version is 18 or newer, skip this step.

💡 You do NOT need to run npm install. The DVPX reflector has no external dependencies; it uses Node's built-in modules only.

2) Get the files from the repository

The files live in a public GitHub repository. Using Git is recommended: updating later becomes a single command.

Do not leave out the /opt/dvpx-reflector path at the end of the command: giving the target directory explicitly puts the files exactly there (no mkdir needed, git creates the directory itself). We pick that path on purpose — the dvpx-reflector.service file shipped in the repository uses the same path, so you can copy it without editing. The repository is the reflector only; the dashboard runs at the centre of the network and is never installed on your server.

3) config.json

Save the template below as /opt/dvpx-reflector/config.json. The dashboard URL is pre-filled for this installation; replace the token with the value the administrator gave you. The ports must MATCH what you wrote in your request.

4) Firewall

Your provider may run a SECOND firewall in front of the server (Hetzner, AWS, Oracle Cloud, Azure, Google Cloud…). This is the most commonly missed step: add the same rules there too.

5) Trial run

If the log looks clean (registered with the dashboard, TCP/UDP listening), stop it with Ctrl+C and move on to the service.

6) Install as a service

This keeps it running after you close the SSH window, starts it on boot and restarts it after a crash. Do NOT run the reflector as root.

If your install path is not /opt/dvpx-reflector, fix the WorkingDirectory, Documentation and ReadWritePaths lines in the service file. If node lives elsewhere (which node), fix ExecStart too.

1) Back up first

config.json is your own file and is not kept in the repository, so git pull never overwrites it. Still, a backup costs a minute and saves you from having to ask for a new token.

2) Download and install the update

If you installed with Git (the recommended path) it is five commands in total. The first line installs git if it is missing.

If you did not install with Git (one-off migration)

If the directory is not a git repository (you installed from a zip), re-clone it once; your config.json is preserved. Every later update becomes the five commands above.

3) How do you know it worked?

You will see the new version number in the log. The moment the reflector checks in with the dashboard (a few minutes at most) the red warning on your panel clears ITSELF — there is no “I read it” button and you need not notify anyone. If the warning stays, the update did not actually take effect.

After every config.json change you must RESTART the service; the file is read only at startup.

• Keep your reflector up. Announce planned maintenance in advance.
• Keep the software current; move to a new version when the administrator announces one.
• Keep the API token secret. If you suspect it leaked, use "Request New Token" on your panel IMMEDIATELY.
• When something goes wrong, write from 💬 Chat on your panel; the conversation updates live and every dashboard administrator sees it.
• Do not run the reflector as root; use an unprivileged user.
• Keep your instant-messaging details current — that is how you are reached in an emergency.
• Do not record user traffic or pass it to third parties.

A reflector-operator account is DELIBERATELY very limited. Nothing you do touches the network directly; everything passes through administrator approval.