System Performance (M. Fowler)

Last week I started reading Patterns of Enterprise Application Architecture (Fowler, 2002). In the introduction there is a great examination of how to distinguish between different types of performance requirement.

Response Time - Initial quantity of time taken.
Responsiveness - Quantity of time taken to respond to a work request, but not necessarily to perform the request.
Latency - The time taken to issues commands, e.g. for a RPC between to systems.
Throughput - The volume of data or number of requests that can be processed in a unit of time.
Load - The measure of work being performed, 0 to 100% load for throughput/concurrent sessions.
Load sensitivity - A measure of the Responsiveness/Throughput/Latency degradation of performance when Load is increase or decreased.
Efficiency - The resources consumed to deliver the system. CPU cycles/Memory/Disk/kilowatts of power.
Capacity - Maximum or upper limit of Load or Throughput.
Scalability - A measure of the systems ability to scale by adding more resources, usually hardware.