Understanding the circuit breaker pattern
It's common for a distributed system to make a call to other downstream services, and the call could fail or hang without response. You will often see code that retries the failed call several times. The problem with a remote service is that it could take minutes or even hours to correct, and an immediate retry might end up in another failure. As a result, end users wait longer to get an error response while your code retries several times. This retry function would consume the threads, and it could potentially induce a cascading failure.
The circuit breaker pattern is about understanding the health of downstream dependencies. It detects when those dependencies are unhealthy and implements logic to gracefully fail requests until it detects that they are healthy again. The circuit breaker can be implemented using a persistence layer to monitor healthy and unhealthy requests over the past request interval.
If a defined percentage...