We have restricted our view of bottlenecks to CPU, memory, and disk because these are the common bottlenecks we experience on our systems. We should not forget other bottlenecks such as bottlenecks due to saturated and overloaded networks. We could think of ways of resolving such bottlenecks such as using faster network cards, Automatic Port Aggregation, and segmenting our network into subnets. There are applications bottlenecks relating to particular aspects of how applications function, e.g., the size of global data areas and how they relate to accessing data through the filesystem buffer cache. The scenarios we have looked at so far should keep you busy for some time.
