Analysis method and apparatus

• US 8,560,667 B2
• Filed: 04/25/2007
• Issued: 10/15/2013
• Est. Priority Date: 10/28/2004
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Kikuchi, Shinji, Yokoyama, Ken, Takeyama, Akira, Shimazaki, Kenichi, Maeda, Takamitsu, Ishibashi, Koji, Shindo, Seiya, Tsuro, Koutarou
• Primary Examiner(s): Dailey, Thomas
• Assistant Examiner(s): TRAN, NAM T
• Application Number:
  US11/739,946
• Publication Number:
  US 20070214261A1
• Time in Patent Office: 2,365 Days
• Field of Search: None
• US Class Current: 709/224
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3423 where the assessed time is active or idle time
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 2201/86 Event-based monitoring
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/0852 Delays

Abstract

An analysis method for carrying out an analysis for responses of a computer system including a plurality of servers, includes: obtaining data concerning a CPU utilization ratio of each of the plurality of servers from the computer system, and storing the data concerning the CPU utilization ratio into a CPU utilization ratio storage; obtaining processing history data generated in the computer system, generating data of a request frequency by users of the computer system, and storing the processing history data into a request frequency data storage; and estimating an average delay time in each server by using the CPU utilization ratio of each server, which is stored in the CPU utilization ratio storage, and the request frequency stored in the request frequency data storage, and storing the estimated average delay time into a server delay time storage. By carrying out such a processing, the analysis can be carried out without changing the computer system to be analyzed and any additional cost.

Claims

1. An analysis apparatus for carrying out an analysis for responses of a computer system including a plurality of servers, comprising:

a unit to obtain data concerning a CPU utilization ratio of each of said plurality of servers from said computer system;

a storage device;

a unit to obtain processing history data generated in said computer system, to generate data of a request frequency by users of said computer system, and to store said data of said request frequency into said storage device; and

an estimate unit to estimate an average delay time in each said server by using said CPU utilization ratio of each said server and said request frequency stored in said storage device, and

wherein said estimate unit comprises;

a consumed CPU time estimate unit to estimate an average consumed CPU time per one request for each said server by carrying out a regression analysis using said CPU utilization ratio of each said server and said request frequency; and

a server delay time estimate unit to estimate an average delay time in each said server by using said average consumed CPU time per one request for each said server and said CPU utilization ratio of each said server, and

said consumed CPU time estimate unit calculates inclination of a regression straight line on a plane mapped by said CPU utilization ratio and said request frequency as said average consumed CPU time per one request for each said server.

2. The analysis apparatus as set forth in claim 1, wherein said consumed CPU time estimate unit carries out the regression analysis using said CPU utilization ratio of each said server and said request frequency, which were obtained during a period when said request frequency was greater than a threshold.

3. The analysis apparatus as set forth in claim 1, wherein said server delay time estimate unit reads out a pertinent coefficient value representing a relation between said average consumed CPU time per one request for said server and said average delay time in said server by referring to a matrix storage storing said coefficient values for each predetermined unit of said CPU utilization ratio, which is an element to determine said coefficient value, and for each number of CPUs, and calculates said average delay time in said server from the read coefficient value and said average consumed CPU time per one request for said server.

4. The analysis apparatus as set forth in claim 1, further comprising:

a unit to estimate, when said plurality of servers included in said computer system are categorized according to job types to be executed, said average delay time for each said category.

5. The analysis apparatus as set forth in claim 4, further comprising:

a unit to estimate an average delay time for the entire computer system by using said average delay time in each said server.

6. The analysis apparatus as set forth in claim 5, further comprising:

a unit to obtain an average actual measurement value of a response time for a request by a user; and

a unit to estimate a delay time, which occurs in a portion other than said servers, by a difference between said average actual measurement value and said average delay time of said entire computer system.

7. The analysis apparatus as set forth in claim 6, further comprising:

a unit to calculate, for each said category, a correlation coefficient between a total sum of said average consumed CPU times and said request frequency, and to determine a confidence degree of said average delay time for each said category based on said correlation coefficient; and

a correction unit to correct said average delay time for each said category based on said confidence degree of said average delay time for each said category.

8. The analysis apparatus as set forth in claim 7, wherein said correction unit comprises:

a unit to sort said average delay times in descending order of said confidence degree;

a unit to accumulate said average delay times for each said category in said descending order of said confidence degree, and to identify an order of said confidence degree at which the accumulated average delay time becomes a maximum value less than said delay actual measurement value; and

a unit to correct a delay time in a next order of the identified order of said confidence degree to a difference between said delay actual measurement value and a value obtained by accumulating said average delay times for each said category in said descending order of said confidence degree up to said identified order of said confidence degree.

9. The analysis apparatus as set forth in claim 1, further comprising:

a unit to change, when said request frequency is changed, said CPU utilization ratio of each said server according to the changed request frequency;

a unit to estimate an average delay time for each said server by using the changed CPU utilization ratio for each said server; and

a unit to output said average delay times for each said server after and before the change in a comparable manner.

10. The analysis apparatus as set forth in claim 1, further comprising:

a unit to change, when a number of CPUs is changed, said CPU utilization ratio of each said server according to the changed number of CPUs;

a unit to estimate an average delay time in each said server by using the changed CPU utilization ratio of each said server and the changed number of CPUs; and

a unit to output said average delay times of each said server before and after the change, in a comparable manner.

11. The analysis apparatus as set forth in claim 1, further comprising:

a unit to calculate, when a number of servers is changed, calculating an average consumed CPU time per one request for each said server according to the changed number of servers;

a unit to calculate a CPU utilization ratio for each said server after the change by using a number of CPUs and said average consumed CPU time per one request for each said server after the change; and

a unit to estimate an average delay time for each said server after the change by using said average consumed CPU time per one request for each said server after the change and said CPU utilization ratio for each said server after the change.

12. The analysis apparatus as set forth in claim 11, further comprising:

a unit to estimate an average delay time for each category defined by classifying said plurality of servers in said computer system according to a job type to be executed by using said average delay time for each said server after the change and the changed number of servers.

13. A non-transitory computer-readable storage medium storing a program for causing a computer to execute a process of an analysis for responses of a computer system including a plurality of servers, said process comprising:

obtaining data concerning a CPU utilization ratio of each of said plurality of servers from said computer system;

obtaining processing history data generated in said computer system, and generating data of a request frequency by users of said computer system; and

first estimating an average delay time in each said server by using said CPU utilization ratio of each said server and said request frequency, and

wherein said first estimating comprises;

second estimating an average consumed CPU time per one request for each said server by carrying out a regression analysis using said CPU utilization ratio of each said server and said request frequency; and

third estimating an average delay time in each said server by using said average consumed CPU time per one request for each said server and said CPU utilization ratio of each said server, and

said second estimating comprises calculating inclination of a regression straight line on a plane mapped by said CPU utilization ratio and said request frequency as said average consumed CPU time per one request for each said server.

14. An analysis method for carrying out an analysis for responses of a computer system including a plurality of servers, said method comprising:

obtaining, by a computer, data concerning a CPU utilization ratio of each of said plurality of servers from said computer system;

obtaining, by said computer, processing history data generated in said computer system, generating data of a request frequency by users of said computer system, and storing said data of said request frequency into said storage device; and

first estimating, by said computer, an average delay time in each said server by using said CPU utilization ratio of each said server and said request frequency stored in said storage device,

wherein said first estimating comprises;

second estimating an average consumed CPU time per one request for each said server by carrying out a regression analysis using said CPU utilization ratio of each said server and said request frequency; and

third estimating an average delay time in each said server by using said average consumed CPU time per one request for each said server and said CPU utilization ratio of each said server, and

said second estimating comprises calculating inclination of a regression straight line on a plane mapped by said CPU utilization ratio and said request frequency as said average consumed CPU time per one request for each said server.