We often hear the terms QPS/TPS in our daily work, and we are often asked how much your system throughput is. From a business point of view, this problem can be understood as the maximum number of user visits per second that the application system can accept. Or the maximum number of requests that can be processed per second;
QPS: The number of times the request is processed per second; note that it is processed here. Specifically, it refers to sending a request to the server and successfully returning the result after processing. It can be understood that there is a counter in the server. Each time a request is processed, 1 is added. After 1 second, counter=QPS.
TPS: The number of transactions processed per second. Generally, TPS is for the entire system. How many transaction processing can an application system complete in 1s, a transaction may correspond to multiple requests in distributed processing, and QPS is used more to measure the processing capability of a single interface service.
Concurrency: The number of requests that the system can handle at the same time
RT: Response time, the average processing time required to process a request
Calculate the relationship:
QPS = Concurrency / Average Response Time
Concurrency = QPS * average response time