Live Stream Failover
When your primary feed drops, Streamrun switches to a backup automatically. No manual intervention, no scrambling mid-broadcast. Your outgoing stream that stays live.
Why failover matters
When you go live, downtime is not just inconvenient. It means lost viewers, lost engagement, and potentially lost revenue. Whether you are broadcasting a major event to tens of thousands or a private stream to a smaller audience, viewers expect the feed to stay online. Failover is the safety net that makes that possible.
A dropped connection, encoder crash, or network glitch takes the whole stream offline. Viewers see a buffering screen. You find out minutes later.
Problems are detected in real time. Streamrun switches to a backup source automatically, keeping the stream live while you resolve the issue.
Redundant inputs
Reliable failover starts with redundancy. Streamrun accepts multiple input streams so that a single failure never takes your broadcast offline.
Active / Standby
Primary encoder
SRT / RTMP — active
Backup encoder
SRT / RTMP — on standby
Streamrun
Twitch
Uninterrupted output
YouTube
Uninterrupted output
One input is live while a backup is ready. If the primary drops, Streamrun activates the standby automatically. Simple to configure and reliable for most setups.
Active / Active
Input A
SRT — sending continuously
Input B
SRT — sending continuously
Streamrun
Best available signal
Forwarded to all destinations
Both inputs stream simultaneously. Streamrun scores each one in real time and always forwards whichever is healthier. The tradeoff is higher bandwidth usage — the payoff is the strongest possible protection against unstable network conditions.
Fast detection
Switching fast means detecting problems fast. Streamrun watches multiple stream health signals continuously and acts the moment something goes wrong.
Connection drop
Detected the moment the TCP or SRT session closes. No grace period needed — a missing connection is an unambiguous signal.
Packet loss & bitrate
Rising packet loss or a sustained bitrate drop indicates a degrading connection before it fails completely, enabling proactive switching.
Frozen or missing video
A connected encoder can still deliver a frozen frame or no video at all. Streamrun detects stalled video separately from connection health.
SRT network statistics
SRT exposes real-time latency, round-trip time, and retransmission data. This enables faster and more accurate detection than TCP-based protocols.
In well-configured setups, failover happens fast enough that viewers experience only a brief glitch — or no visible interruption at all.
Implementation patterns
Failover can be layered at multiple levels depending on what you are protecting against.
Gateway-level failover (recommended)
Streamrun handles everything automatically at the cloud level. Configure primary and backup inputs, define detection thresholds, and let Streamrun do the rest. No additional hardware or custom logic required.
- •Add a primary and a backup input to your pipeline in the Streamrun editor.
- •Streamrun monitors both inputs and switches automatically on failure.
- •Events are logged for post-broadcast analysis.
- •The primary feed is restored automatically once it is stable again.
Encoder redundancy
Run two independent encoders sending the same content to Streamrun as separate inputs. Protects against encoder crashes, overheating, and device failure.
Geographic redundancy
For critical broadcasts, route your stream through separate network paths or from different physical locations. Combined with Streamrun's cloud infrastructure, this guards against ISP outages and regional network issues.
Visibility and alerts
Failover should never happen silently. Streamrun keeps you informed so you know what is happening in real time.
Input status dashboard
See the health of every input at a glance — bitrate, packet loss, and connection state updated in real time.
Failover notifications
Receive an alert the moment Streamrun activates a backup input, so you can investigate and resolve the underlying issue immediately.
Event log
Every failover event is recorded with a timestamp and the reason it triggered. Useful for post-broadcast troubleshooting and improving your setup over time.
Test before you go live
A failover setup that has not been tested should not be trusted for mission-critical broadcasts. Run these checks during a rehearsal, before the real event.
Disconnect the primary network mid-rehearsal
Confirm the backup activates within your target window.
Simulate packet loss or bandwidth throttling
Verify detection thresholds trigger at the right level.
Kill the encoder process unexpectedly
Check that the backup input takes over without manual action.
Restore the primary and confirm auto-recovery
Ensure Streamrun switches back without manual intervention.
Keep your stream live, no matter what
Free 14-day trial. No credit card required.