observedthroughput

Observed throughput is the actual rate at which data is successfully transferred over a network or system during a measurement interval. It is an empirical metric that reflects real-world performance, as opposed to theoretical capacity or maximum data rate.

Measured in bits per second or bytes per second, observed throughput can be end-to-end or per-link and is typically reported over a specified time window. It can vary with traffic patterns, routing, and conditions, and may include or exclude protocol overhead depending on the measurement method.

Several factors can reduce observed throughput below link capacity: protocol overhead, retransmissions due to errors, congestion control and queueing delays, cross-traffic, and, in wireless environments, interference and backoff. The measurement method itself also influences the result, as active tests may capture peak or average rates during probing while passive monitoring reflects ongoing real flows.

Observed throughput is related to, but distinct from, theoretical throughput. It is often contrasted with goodput, the usable data delivered to applications, which excludes or accounts for overhead differently depending on definition.

Measurement approaches include active tests, such as speed tests or iPerf, and passive monitoring of live traffic. Reporting is usually over a fixed interval to smooth variability and provide comparable results.

Use cases for observed throughput include performance evaluation, capacity planning, SLA verification, and troubleshooting, helping diagnose whether observed performance meets expectations under actual operating conditions.