Source Failover MediaConnect

Question

Hello,

As I am learning about M&E services in AWS, I came across the concept of Source Failover in MediaConnect. Source Failover aims to provide redundancy for live media streams.

Merge mode combines the sources into a single stream, allowing a graceful recovery from any single-source loss. If you set the failover mode to Merge, you can set the recovery window, which is the size of the buffer (delay) that you want MediaConnect to maintain. A larger recovery window means a longer delay in transmitting the stream, but more room for error correction. A smaller recovery window means a shorter delay, but less room for error correction. Sources used this way need to be binary identical, which means that they need to have originated from the same encoder. MediaConnect must also receive content from the two sources at the same time.

https://docs.aws.amazon.com/mediaconnect/latest/ug/source-failover.html

How can we provide redundancy if the source comes from the same encoder? What happens if one of the on-prem encoders or other type of encoders fails?

Accepted Answer

Hello,

For Merge Mode, your encoder will indeed be a single point of failure. However, you can send each of the streams out of 2 different network interfaces, and have each network interface attached to a different ISP, so that the routes into MediaConnect are unique and more resilient. 
If you send 2 streams from 2 separate encoders, MediaConnect can still perform input failover, but it's not completely hitless like the Merge feature is.
With Merge what would happen if the upstream encoder failed, is that your stream would go down.

-Dirk

Source Failover MediaConnect

Relevant content