Apparatus, method and product for optimizing software system workload performance scenarios using multiple criteria decision making
First Claim
1. A method of evaluating performance of at least one software system, the method comprising:
- using a computing device, identifying a plurality of performance criterion using a plurality of operational laws, wherein a plurality of operational laws comprise a Little Law equation and a utilization law, wherein a utilization law is device throughput multiplied by service demand, whereby service demand for the devices remain constant;
gathering from the user a set of application details, which are associated with an application scenario;
mapping the application details with a patterning module that uses a knowledge base to rank the criteria to reflect the requirements of the application scenarios based on context details of user input; and
using a computing device, selecting a plurality of performance scenarios for the at least one software system based on at least one of the plurality of performance criterion using at least one Multiple Criteria Decision Making (MCDM) technique;
wherein the MCDM technique comprises computing a plurality of ranks of the application scenarios for each of a cost criteria and a benefit criteria individually, aggregating the application scenario ranks for the benefit criteria and cost criteria, and computing an overall criticality rank for the application scenarios by taking the ratio of the benefit criteria and cost criteria ranks.
2 Assignments
0 Petitions
Accused Products
Abstract
The present technique is an apparatus and method evaluating software performance. The method identifies performance scenarios using a knowledge base and selects the performance scenarios from a context module using operational laws. The system analyzes performance criticality of an application workload. Furthermore, the system comprises a context module that ranks the performance criterions depending on the input and the knowledge base stores and retrieves the performance criterions using the operational laws. The knowledge base comprises the criterions and their relevant ranks based of the application context. The present technique automates classification of performance criterions into benefit and cost categories with the usage of the operational laws.
35 Citations
28 Claims
-
1. A method of evaluating performance of at least one software system, the method comprising:
-
using a computing device, identifying a plurality of performance criterion using a plurality of operational laws, wherein a plurality of operational laws comprise a Little Law equation and a utilization law, wherein a utilization law is device throughput multiplied by service demand, whereby service demand for the devices remain constant; gathering from the user a set of application details, which are associated with an application scenario; mapping the application details with a patterning module that uses a knowledge base to rank the criteria to reflect the requirements of the application scenarios based on context details of user input; and using a computing device, selecting a plurality of performance scenarios for the at least one software system based on at least one of the plurality of performance criterion using at least one Multiple Criteria Decision Making (MCDM) technique; wherein the MCDM technique comprises computing a plurality of ranks of the application scenarios for each of a cost criteria and a benefit criteria individually, aggregating the application scenario ranks for the benefit criteria and cost criteria, and computing an overall criticality rank for the application scenarios by taking the ratio of the benefit criteria and cost criteria ranks. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A system for evaluating performance of at least one software system, the system comprising:
-
a computer processor adapted to allow modules to function in a system; an input module adapted to obtain a plurality of performance criterion depending on at least one input and storing into a knowledge base using a plurality of operational laws, wherein a plurality of operational laws comprise a Little Law equation and a utilization law, wherein a utilization law is device throughput multiplied by service demand, whereby service demand for the devices remain constant; a gathering module adapted to gather from the user a set of application details, which are associated with an application scenario; a patterning module that uses the knowledge base to rank the criteria to reflect the requirements of the application scenarios based on context details of user input; a selection module adapted to select a plurality of performance scenarios for the at least one software system depending on at least one of the plurality of performance criterion using at least one Multiple Criteria Decision Making (MCDM) technique; and an output module adapted to display the plurality of performance scenarios based on the at least one input and the plurality of performance criterion and at least one rank using the at least one MCDM technique, wherein the MCDM technique comprises computing a plurality of ranks of the application scenarios for each of a cost criteria and a benefit criteria individually, aggregating the application scenario ranks for the benefit criteria and cost criteria, and computing an overall criticality rank for the application scenarios by taking the ratio of the benefit criteria and cost criteria ranks. - View Dependent Claims (19, 20, 21, 22, 23, 24)
-
-
25. A computer program product tangibly embodying a plurality of instructions, which when executed by a computer processor cause said processor to carry out steps for evaluating performance of at least one software system, the computer program product comprising:
-
program code adapted to select a plurality of performance criterion using a plurality of operational laws, wherein a plurality of operational laws comprise a Little Law equation and a utilization law, wherein a utilization law is device throughput multiplied by service demand, whereby service demand for the devices remain constant; program code adapted to gather from the user a set of application details, which are associated with an application scenario program code adapted to create and represent a knowledge base indicative of the plurality of performance criterion using the plurality of operational laws; program code adapted to map the plurality of performance criterion to the knowledge base to reflect the requirements of the application scenario based on context details of user input; and program code adapted to rank a plurality of scenarios of the at least one software system depending on the mapping of the plurality of performance criterion using the at least one Multiple Criteria Decision Making (MCDM) technique wherein the MCDM technique comprises computing a plurality of ranks of the application scenarios for each of a cost criteria and a benefit criteria individually, aggregating the application scenario ranks for the benefit criteria and cost criteria, and computing an overall criticality rank for the application scenarios by taking the ratio of the benefit criteria and cost criteria ranks. - View Dependent Claims (26, 27, 28)
-
Specification