| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To achieve the valid use of a CPU resource by equally distributing the analysis processing of a message to each CPU. <P>SOLUTION: A storage part 16 is configured to store the history information of the analysis processing time and data quantity of a response message for every request classification. A calculation part 11 is configured to, when receiving a packet, calculate the relational expression of the analysis processing time and data quantity of the response messages of the same request classification by referring to the history information. A data quantity estimation part 12 is configured to estimate the data quantity of the response message from the latest statistics of the data quantity. A prediction part 13 is configured to predict the processing time by the relational expression from the estimation value of the data quantity. A distribution part 14 is configured to distribute the analysis processing of data to a processor, namely, a control part 17 so that the analysis processing time can be equalized on the basis of the predicted processing time. <P>COPYRIGHT: (C)2013,JPO&INPIT</p> |
