| 摘要 |
Measuring performance and capacity of a networked server, the method including: initializing each client machine of a cluster of client machines with a fixed number of client applications, wherein each client application uses a different type of operations to communicate with a server; generating a first feedback process of a current state of each client machine and overall behavior of the client applications; enabling each client application to self adjust its own operation based on the first feedback process; and setting up a second feedback process for the server in which the server and the cluster of client machines reach a balance point of a client count for the server.
|
