Checking regional health

At this point we can justify why we run our serverless systems in multiple regions and we appreciate that we have been preparing for this all along. However, before we go any further, we need to know how to check the health of the cloud regions we are using, because it doesn't help to run in multiple regions if we cannot identify when it is time to failover from an unhealthy region.Like so many other things, serverless is different when it comes to health checks. At first glance this may seem like a familiar exercise, but regional health checks are different from traditional health checks. They have a lot in common, but regional health checks work at a higher order. Let's review how traditional health checks work and then use that as juxtaposition for our regional health checks.

Traditional health checks

In a traditional, non-serverless system we need to run clusters of machine instances across multiple availability zones, and we place a load balancer...