High Availability Layers
The term “High Availability” is used generically these days and has a different meaning to different people. Per above diagram, each layer requires its own High Availability for redundancy. For example, at the Physical layer providing multiple NIC cards, Switches, etc., is great for eliminating single point of failure on the physical layer but it does not address failures at the above layers, Virtualization, OS or/and Application layer. On the other hand, redundancy at application level does not make the underlying physical hardware or other layers redundant. Each layer requires its own High Availability planning and it is important to clarify when a customer says I want High Availability.