Reducing internal retention time of processing requests on a web system having different types of data processing structures

A method for reducing internal retention time of processing requests on a web system, comprising:

• an extraction step of extracting, from operation data of the web system having the capacity control parameters including a maximum queue length indicating a maximum size of a queue which temporarily stores the processing requests and a maximum number of simultaneous executions indicating the maximum number of the processing requests that can be processed in parallel, an internal retention time required for the web system to process data for each of different numbers of internal retention requests wherein the number of internal retention requests is a sum of the queue length and the number of simultaneous executions, and wherein the internal retention time includes only processing time of the processing requests that is controlled by the capacity control parameters;
• a coefficient calculation step of applying an approximate equation to the internal retention time for each of the different number of internal retention requests extracted in the extraction step to calculate coefficients of the approximate equation by regression analysis, the approximate equation corresponding to a particular data processing structure of the web system;
• a parameter calculation step of calculating new capacity control parameters including the maximum number of simultaneous executions and the maximum queue length, at which the internal retention time at different amounts of load of the web system is minimized, from the approximate equation having the coefficients calculated in the coefficient calculation step;
• and
• a transmission step of transmitting the new capacity control parameters calculated by the parameter calculation step to the web system, wherein the web system executes the processing requests with the new capacity control parameters.

A computer program product stored on a non-transitory computer-readable medium for reducing internal retention time of processing requests on a web system, which when executed by a processor causes the processor to:

• extract from operation data of the web system having capacity control parameters including a maximum queue length indicating a maximum size of a queue which temporarily stores the processing requests and a maximum number of simultaneous executions indicating the maximum number of the processing requests that can be processed in parallel, an internal retention time required for the web system to process data for each of different numbers of internal retention requests wherein the number of internal retention requests is a sum of the queue length and the number of simultaneous executions, and wherein the internal retention time includes only processing time of the processing requests that is controlled by the capacity control parameters;
• apply an approximate equation to the internal retention time for each of the different number of internal retention requests extracted from the operation data to calculate coefficients of the approximate equation by regression analysis, the approximate equation corresponding to a particular data processing structure of the web system;
• calculate new capacity control parameters including the maximum number of simultaneous executions and the maximum queue length, at which the internal retention time at different amounts of load of the web system is minimized, from the approximate equation having the calculated coefficients;
• and
• transmit the new capacity control parameters to the web system, wherein the web system executes the processing requests with the new capacity control parameters.