In practice, this result is achieved by providing backup solutions for the various components that contribute to the functioning of the overall infrastructure. If one of them stops working, the secondary solution would immediately take over automatically, allowing the continuity of the service without causing any disturbance to users – who would probably not even notice the change. This mechanism is called failover, as the interrupted process “falls back” to the designated replacement component.
It should be noted that normally it is not the entire architecture that is replaced, but only the single malfunctioning component. To understand how this happens, let's examine a representative example of a very common case: an e-commerce platform in HA. In this type of application we would find different levels of articulation.
