Replicating across regions
It is inevitable that a given cloud provider will experience a news-worthy regional disruption. It is not a matter of if but of when. In my experience, this happens about every 2 to 3 years or so. When such an event does occur, many systems have no recourse and become unavailable during the disruption, because they are only designed to work across multiple availability zones within a single region. Meanwhile, other systems barely experience a blip in availability because they have been designed to run across multiple regions.
In Chapter 3, Taming the Presentation Tier, we saw how session consistency and offline-first mechanisms help the presentation layer ensure a seamless transition during a regional disruption. In Chapter 6, A Best Friend for the Frontend, we will cover the regional health check and regional failover mechanisms that are leveraged by BFF services. In Chapter 4, Trusting Facts and Eventual Consistency, we covered the implications of multi...