Testing failover and reconnect safely

Failover and reconnect placeholders are operational safety features. Test them before an event, not for the first time during one.

Use a private or low-risk channel, start the primary source, confirm the public page, then stop the source for a short period. Confirm the expected backup or reconnect slate appears, restore the source and check that the primary returns. Record the timestamps so you can compare the ingest and destination logs.

A placeholder should never show stream keys, private URLs or internal error details. Failover cannot fix a revoked credential, an unavailable destination or a bad source format. If the test does not converge, disable the feature for the event and use a known-good single-source workflow while investigating.