Most of the software architectures in use today are very complex. We need to make the architecture of a software application as simple as possible so that the process of interaction can be made as simple as possible. This will make it possible even for non-computer experts to interact with these applications easily. One way to achieve this is by use of microservices. In microservices, a software application needs to be broken down into a set of individual components called services. One also needs to determine whether the service in the same application should be made to communicate with each other or not. Communication between services can be done either synchronously or asynchronously. One should also know how to divide or decompose into services. There are various strategies on how this can be done. One also needs to understand how to register the instances of a particular service with the service registry. Registration is done on starting up. After shutdown of the application, deregistration occurs, where the instances of the service are unregistered from the service registry. One can have the services register themselves in what we call self-registration or can choose to personally register the services. There exists various microservices architecture patterns. Each has its own advantages and disadvantages.
The following topics have been discussed:

Monolithic Architecture
API Gateway Microservices Architecture
Client-side service discovery
Server-side Service Discovery
3rd Party Registration
Self Registration
Service Registry
Single service instances per host
Multiple service Instances per Host
Service Instance per VM
Service instance per container

Microservices Architecture: Make the architecture of a software as simple as possible - heftet, Engelsk, 2016