发明名称 Apparatus and method for predicting a processing time of a computer
摘要 Load information and a first processing time are provided in association with each of a plurality of first time segments that each have a fixed duration time and are included in a first time period, where the load information indicates a load condition of a target computer that executed a target job during the each first time segment, and the first processing time indicates a running time of the target job during the each first time segment. One or more first time segments each having a predetermined analogous relationship with a second time segment in a second time period during which the target job is expected to be executed by the target computer are selected to predict a second processing time indicating a running time of the target job during the second time period based on the first processing times associated with the selected one or more first time segments.
申请公布号 US8793693(B2) 申请公布日期 2014.07.29
申请号 US201213428029 申请日期 2012.03.23
申请人 Fujitsu Limited 发明人 Inagaki Takahiro
分类号 G06F9/46 主分类号 G06F9/46
代理机构 Staas & Halsey LLP 代理人 Staas & Halsey LLP
主权项 1. A computer-readable, non-transitory medium storing instructions for allowing a computer system to execute a procedure comprising: receiving load information and a first processing time in association with each of a plurality of first time segments that each have a fixed duration time and are included in a first time period, the load information indicating a load condition of a target computer that executed a target job during the each of the plurality of first time segments, the first processing time indicating a running time of the target job during the each of the plurality of first time segments; selecting from among the plurality of first time segments, one or more first time segments each having a predetermined analogous relationship with a second time segment included in a second time period during which the target job is expected to be executed by the target computer, based on the load information associated with the plurality of first time segments; and predicting a second processing time indicating a running time of the target job that is expected to be executed during the second time period, based on the first processing time associated with the selected one or more first time segments, wherein the load information includes an average value of load amounts that were imposed on the target computer during each of the plurality of first time segments; and it is determined that one of the plurality of time segments in the first time period has the predetermined analogous relationship with the second time segment, based on a difference between an average value of load amounts imposed on the target computer during the one of the plurality of first time segments and an average value of load amounts imposed on the target computer during the second time segment in the second time period.
地址 Kawasaki JP