Every performance characteristic of a system belongs to one of two Performance Views:
The performance views are complementary, i.e. metrics from both views are required to understand and improve a system's performance.
External view of performance characteristics
External performance views are a system's performance characteristics as observed and measured from outside the tested system.
External performance figures show the aggregated performance efforts of a system.
External performance metrics are also measurements of system's service delivery limits. Examples of external view performance characteristics are response times, response variations, different capacity metrics, availability, stability, and many more.
External View performance characteristics show user esperienced performance, but doesn't explain the system's service delivery limits.
Internal view of performance characteristics
Internal performance views are system's performance characteristics as observed and measured from inside the tested system.
Internal performance figures show the performance of a system's components.
Performance characteristics of a system's components are measurements of what sets the limits the system's external performance. Internal performance measurements are made to find internal sources of deviations from performance requirements such as capacity bottlenecks.
Internal performance metrics are measurements of system resource usage and other factors that explain the external performance metrics. Examples of Internal View performance characteristics are CPU load, memory usage, queue lengths, and many more.
Internal View performance characteristics show the cause a system's service delivery limits, but doesn't explain the user esperienced performance.