Estimating service resource consumption based on response time

Patent metadata

Field	Value
Publication number	US-9804943-B2
Application number	US-58090109-A
Country	US
Kind code	B2
Filing date	Oct 16, 2009
Priority date	Oct 16, 2009
Publication date	Oct 31, 2017
Grant date	Oct 31, 2017

Abstract

Implementations of the present disclosure provide computer-implemented methods including defining a workload comprising a plurality of service requests, each service request corresponding to a class of a plurality of classes, applying the workload to a computer system that receives and processes service requests, measuring a response time of the computer system for each request of the workload, estimating a mean service demand for each class based on the response times and a base queuing model that represents the computer system, and generating the queuing model based on the mean service demands and characteristics of the workload.

First claim

What is claimed is:

1. A computer-implemented method of evaluating a performance of a computer system based on a queueing model, comprising:
defining a workload comprising a plurality of service requests, each service request corresponding to a class of a plurality of classes;
applying, by one or more processors, the workload to the computer system that receives and processes the plurality of service requests;
measuring, by the one or more processors, a response time of the computer system for each request of the workload to provide a plurality of measured response times;
determining, by the one or more processors, a mean response time for each class based on the plurality of measured response times;
estimating, by the one or more processors, a mean service demand for each class based on the mean response time for a respective class and a base queueing model that represents the computer system to provide a plurality of mean service demands, each of the plurality of mean service demands defining an average time a respective service request is attended to by the computer system;
generating, by the one or more processors, the queueing model based on the mean service demands and characteristics of the workload, the queueing model modeling queuing of the plurality of service requests submitted to the computer system; and
processing the queueing model using a plurality of inputs to evaluate the performance of the computer system.

2. The method of claim 1 , further comprising determining a plurality of arrival queue-lengths corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of arrival queue-lengths.

3. The method of claim 2 , wherein each arrival queue-length of the plurality of arrival queue-lengths is determined from log files that report a time of arrival and departure of requests.

4. The method of claim 1 , further comprising determining a plurality of residual times corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of residual times.

5. The method of claim 1 , wherein estimating a mean service demand comprises estimating mean service demands in the base queueing model using one of linear regression and maximum likelihood method analyses based on the measured response times.

6. The method of claim 1 , wherein generating the queueing model comprises parameterizing the base queueing model using the mean service demands.

7. A non-transitory computer-readable storage medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for evaluating a performance of a computer system based on a queueing model, the operations comprising:
defining a workload comprising a plurality of service requests, each service request corresponding to a class of a plurality of classes;
applying the workload to a computer system that receives and processes the plurality of service requests;
measuring a response time of the computer system for each request of the workload to provide a plurality of measured response times;
determining a mean response time for each class based on the plurality of measured response times;
estimating a mean service demand for each class based on the mean response time for a respective class and a base queueing model that represents the computer system to provide a plurality of mean service demands, each of the plurality of mean service demands defining an average time a respective service request is attended to by the computer system;
generating a queueing model based on the mean service demands and characteristics of the workload, the queueing model modeling queuing of requests submitted to the computer system; and
processing the queueing model using a plurality of inputs to evaluate the performance of the computer system.

8. The storage medium of claim 7 , wherein the operations further comprise determining a plurality of arrival queue-lengths corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of arrival queue-lengths.

9. The storage medium of claim 8 , wherein each arrival queue-length of the plurality of arrival queue-lengths is determined from log files that report a time of arrival and departure of requests.

10. The storage medium of claim 7 , wherein the operations further comprise determining a plurality of residual times corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of residual times.

11. The storage medium of claim 7 , wherein estimating a mean service demand comprises estimating mean service demands in the base queueing model using one of linear regression and maximum likelihood method analyses based on the measured response times.

12. The storage medium of claim 7 , wherein generating the queueing model comprises parameterizing the base queueing model using the mean service demands.

13. The storage medium of claim 7 , further comprising evaluating a performance of a computer system by processing the queueing model using a plurality of inputs.

14. A system comprising:
a computer system that receives and processes a plurality of service requests;
one or more processors; and
a computer-readable storage medium coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for evaluating a performance of a computer system based on a queueing model, the operations comprising:
defining a workload comprising the plurality of service requests, each service request corresponding to a class of a plurality of classes;
applying the workload to the computer system;
measuring a response time of the computer system for each request of the workload to provide a plurality of measured response times;
determining a mean response time for each class based on the plurality of measured response times;
estimating a mean service demand for each class based on the mean response time for a respective class and a base queueing model that represents the computer system to provide a plurality of mean service demands, each of the plurality of mean service demands defining an average time a respective service request is attended to by the computer system;
generating the queueing model based on the mean service demands and characteristics of the workload, a queueing model modeling queuing of the plurality of service requests submitted to the computer system; and
processing the queueing model using a plurality of inputs to evaluate the performance of the computer system.

15. The system of claim 14 , wherein the operations further comprise determining a plurality of arrival queue-lengths corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of arrival queue-lengths.

16. The system of claim 15 , wherein each arrival queue-length of the plurality of arrival queue-lengths is determined from log files that report a time of arrival and departure of requests.

17. The system of claim 14 , wherein the operations further comprise determining a plurality of residual times corresponding to each request of the workload, wherein estimating a mean service demand is further based on the plurality of residual times.

18. The system of claim 14 , wherein estimating a mean service dema

Assignees

• Kraft Stephan
• Pacheco-Sanchez Sergio
• Casale Giuliano
• Dawson Stephen
• Sap Se

Inventors

• Kraft Stephan
• Pacheco-Sanchez Sergio
• Casale Giuliano
• Dawson Stephen

Classifications

• G06F11/3419

  by assessing time · CPC title

• G06F11/3447 Primary

  Performance evaluation by modeling · CPC title

• G06F11/3461

  Trace driven simulation · CPC title