Is the event replication durable for the eventBrdige global endpoint?

Question

I don't see any statement saying this.

I guess the event replication between two event buses is asynchronous.
If Event A has been put into the primary region, but the Event Bridge in the primary region is disrupted before a successful replication, event A is missing in the secondary region.

If so, is the durability just an  at-most-once semantic?

Accepted Answer

The replication of events between primary and secondary regions is explicitly designed to ensure application resilience during regional disruptions. However, this process is asynchronous, meaning events are not guaranteed to be processed simultaneously in both regions. This implies that during failover, there might be a window where events are not replicated to the secondary region, potentially leading to an "at-most-once" delivery scenario. The Recovery Time Objective (RTO) and Recovery Point Objective (RPO) are crucial metrics here, with AWS aiming for these to be around 360 to 420 seconds, indicating the designed recovery efficiency and the extent of data that could be left unprocessed during a failure.

For ref:

https://docs.aws.amazon.com/eventbridge/latest/userguide/eb-global-endpoints.html

If this has answered your question or was helpful, accepting the answer would be greatly appreciated. Thank you!

For ref:
* https://docs.aws.amazon.com/eventbridge/latest/userguide/eb-global-endpoints.html

If this has answered your question or was helpful, accepting the answer would be greatly appreciated. Thank you!

Is the event replication durable for the eventBrdige global endpoint?

Relevant content