Before your first live event

Use this page in the days before your first real event. A working preview is only the start. Check the full path with a real feed and a viewer outside your network, and write down who checked what.

On this page

Start with a small, private rehearsal

Use a separate test stream and video you have permission to transmit. Tell everyone involved when the test starts and ends. Do not interrupt a production event to test recovery. If this is your first installation, finish the quick start first.

Agree on the actual devices, networks, video quality and audience size you need to serve. A successful local test or a screenshot cannot establish capacity for that audience. Use the sizing guide and your own load test with headroom.

Keep evidence for each check

Check Expected result Save for your team
Encoder and source The named stream is live; the moving picture and expected sound arrive. Encoder settings, source format and test time. Keep the stream key private.
Outside viewer The intended browser opens the watch link. HLS works as a baseline; test WebRTC separately where required. Device, browser version, network, start time and any failures.
Viewer access An allowed viewer can play; a deliberately excluded test viewer is refused. The rule tested and result. Do not save a usable private-link token.
Recording A short test recording opens from the beginning and near the end after publishing stops. Expected duration, actual usable content and any gaps.
Storage and network Free space covers the event, processing and reserve; the tested delivery path has room for the planned load. Bitrate, duration, number of recordings, free space and test method.
Recovery The team detects a controlled test-source interruption and follows the agreed recovery steps. Time to notice, time to restore playback and who acts.
Licence and support The licence covers the event dates, features and channel count; the responsible operator knows the support route. Expiry and grace dates, owner and contact plan. Never share the licence key.

Investigate the first failed check

If the encoder is not connected, changing the website player is unlikely to fix it. If the source is live but only one phone fails, compare that device's browser, codec and network before buying a larger server. Follow the symptom-based checks and change one variable at a time.

Use the deeper tests when needed: time the first frame, check browser playback, budget recording space and run a recovery drill. None of these tests replaces a representative rehearsal of your own deployment.

Leave an operator handoff

  • Name the primary operator and the person who can change the encoder or server.
  • Write down how to recognise a failed stream and which recovery action is allowed.
  • Confirm where the recording and backup will be checked after the event.
  • Record any known limit or failed test that remains. Decide whether to delay the event, change its requirements or complete another rehearsal.

Before an upgrade, make a backup and test the restore on an isolated installation. Do not assume a backup file is usable merely because it exists. Need help? Send a reviewed, redacted problem description through support.

Next: monitoring and everyday care and a detailed event rehearsal.

When have I passed?

When every row you need has an observed result and a name beside it, including a refused viewer, a finished replay and a deliberate source cut, and someone is named who can say go or no-go. One working viewer tells you nothing about a full audience. Keep anything that failed visible, with an action against it, rather than calling the rehearsal a pass.

Follow the build