| 摘要 |
PROBLEM TO BE SOLVED: To calculate the number of necessary servers from the number of processing requests, and to efficiently determine a server capable of being stopped or a start server.SOLUTION: A processing request number maximum value per unit time from clients 100 (#1, #2, ..., #n) is obtained, a processing number per server at a time point is obtained from a stored past processing request number maximum value and the numbers of the servers operating at the time point, the number of the servers necessary to process the present number of the processing requests is calculated from relation with the this-time acquired processing request number maximum value, and it is decided whether or not there is the server capable of being stopped or to be started. When it is decided that there is the server capable of being stopped, it is waited until a request processed by the server is absent without sorting a new processing request from the client 100 (#1, #2, ..., #n) to the target server, and the server is stopped after detecting that the processing request becomes zero. When it is decided that the server to be started is present, the server is immediately started. |
