| 摘要 |
PROBLEM TO BE SOLVED: To previously transfer a job input file and to effectively use the job waiting time by deciding an optimum computer to execute a job before the execution of the job is started. SOLUTION: A user transmits the name of a job, etc., that is used for executing the job to a job queuing server, and this server sends the number of servers which can execute the job, the names of servers, the information on the position of a network to which the servers are connected and the load situations of the servers which can execute the job to a client (S21, S22). The client clears a storage area where the priorities of all servers are stored, calculates the estimated time needed for transferring a job input file to every server from the network position information and produces a list of job executable servers after arranging the servers in the order of higher priorities which are decided in the order of longer estimated necessary times (S23-S25). Then the job input file is transferred to the storage of a server of higher priority and the information necessary for executing the job is sent to the job queuing server (S26, S27).
|
