The VPS streaming checklist
Connect a server you can operate confidently. Work through these checks before enrolling an agent, then keep monitoring separate from broadcast readiness.
1. Start with a supported machine
Use an Ubuntu or Debian Linux VPS that you control. The installer supports amd64 and arm64 machines with systemd. A source build also requires Go; FFmpeg is needed for broadcasting. Choose capacity for your actual source and number of outputs rather than treating a minimum specification as a streaming guarantee.
- Confirm the operating system, CPU architecture, available RAM, and free disk.
- Keep enough disk for working files as well as the system itself.
- Check that you can administer the VPS and inspect its service logs.
2. Check the outbound path
The agent connects outward to your application API. A remote VPS needs a public HTTPS API endpoint with a valid certificate. An address such as 127.0.0.1 or localhost refers to the VPS when used there, not the computer running your development application. The dashboard’s optional public-host field is a saved label; it is not what enables this communication. You do not need to supply an SSH private key or cloud account credentials to register the server.
3. Register, then install
Name the server in the Servers page. Create the short-lived setup file when you are ready to install, transfer it privately with the agent source, and follow the displayed source-build installation steps on the VPS. The installer creates a dedicated service account and a systemd service. Keep the configuration restricted and avoid replacing an existing enrolled identity with another server’s setup file.
4. Read the first heartbeat
Return to Servers and review the most recent heartbeat, system information, and measurement freshness. Unknown values should remain unavailable. An old metric describes an earlier measurement, not the machine’s current condition. If the agent has not reported, inspect its service state and configured API address. A missing heartbeat alone does not identify whether the cause is configuration, networking, or the service itself.
5. Make streaming a separate test
Check detected FFmpeg availability and prepare one short video and one destination. Start a test broadcast, review the output’s process reports, and check the destination platform preview. Stop the test and confirm the stopped state before planning a longer session.
A useful next step.
Read the current product capabilities, or find a guide for the part of your setup you want to improve.