Why Your Website’s Speed Matters: Understanding RT, QPS, and Concurrency

Response Time (RT)

Response Time (RT) is the duration from when a client sends a request until the client receives the server's response. Long RT degrades user experience, so it is a key performance metric that websites aim to minimize.

Using a Disney park analogy, the time a visitor waits in line is measured by recording the start time at the entrance and the end time at the exit, then calculating the difference. This mirrors how RT is calculated: record the timestamp when a request starts and when it ends, and the difference is the RT.

QPS (Queries Per Second)

QPS measures how many requests a system can handle per second. It is a primary indicator of system throughput and performance.

QPS and RT are closely related: increasing QPS often requires reducing RT, but there are many ways to improve QPS beyond simply lowering RT, such as scaling hardware resources.

RT = 并发数 / QPS
QPS = 并发数 / RT

Concurrent Users

Concurrent users refer to the number of users interacting with the server at the same moment. It is not the total number of registered users nor the number of online users over a period; it is the count of active sessions at a specific instant.

For example, the concurrent user count for an e‑commerce product page is the number of users requesting that page simultaneously.

Optimal Thread Count

The optimal thread count is the maximum number of concurrent users a service can effectively handle before performance degrades sharply. Exceeding this threshold leads to resource contention, high CPU load, increased memory usage, and dramatically longer response times.

During performance testing, QPS rises with increasing users until the optimal thread count is reached; beyond that point, QPS plateaus while RT spikes.

Understanding and monitoring RT, QPS, concurrent users, and optimal thread count helps engineers design, tune, and scale web applications for better performance and user experience.