In performance testing, throughput refers to the workload a system can process within a unit of time. It is commonly used to evaluate the data processing capability of networks, servers, or other computing devices within a specific time frame. Throughput can be measured in various units, such as transactions per second (TPS), requests per second (RPS), or data processed per second (e.g., MB/s).
Example:
Assume we are conducting performance testing on a web application whose primary function is processing user-submitted online orders. During the test, we can measure the application's throughput by simulating multiple users submitting orders concurrently. If the application successfully processes 1,800 orders within one minute, we can calculate its throughput as 30 orders per second (1,800 orders / 60 seconds = 30 orders/second).
Why Important:
Throughput is one of the key metrics for measuring system performance, helping us understand the processing capacity the system can achieve in real-world operation. By testing and optimizing throughput, we can ensure the system remains stable under high loads, avoiding service delays or failures due to insufficient processing capacity.
Testing Considerations:
When conducting throughput testing, we need to consider multiple factors such as network latency, server configuration, and concurrent user count, as these can influence the final test results. Therefore, throughput testing should be performed under conditions as close as possible to the production environment to obtain more accurate results. Additionally, we must monitor the system's stability and resource consumption when handling high throughput to ensure it can process high loads without crashing due to excessive resource consumption.