Transaction based workload modeling for effective performance test strategies

• US 8,751,184 B2
• Filed: 07/11/2011
• Issued: 06/10/2014
• Est. Priority Date: 03/31/2011
• Status: Active Grant

• Current Assignee: Infosys Limited
• Original Assignee: Infosys Limited
• Inventors: Gupta, Vikas, Gawande, Amit
• Primary Examiner(s): Lau, Tung S
• Application Number:
  US13/179,647
• Publication Number:
  US 20120253733A1
• Time in Patent Office: 1,065 Days
• Field of Search: 702/120
• US Class Current: 702/120
• CPC Class Codes: G06F 11/3414 Workload generation, e.g. scripts, playback
• CPC Class Codes: G06F 11/3433 for load management allocation of a server based on load conditions G06F9/505; load rebalancing G06F9/5083; redistributing the load in a network by a load balancer H04L67/1029
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/87 Monitoring of transactions

Abstract

A method for creating workload model to test performance of a critical application in a data processing network (112) is disclosed. The method includes receiving (202), at one of the plurality of data processing units (114a), an activity log with a plurality of entities from the data processing network (112). The method further includes creating (204) a transaction log of all transactions conducted using the activity log. Further the method determines (206) a plurality of transaction metrics using the transaction log, wherein the plurality of transaction metrics comprises an average transaction response time, a transaction throughput etc. The method further creates a workload model to test the critical performance of an application in the data processing network (112) based on the plurality of transaction metrics and provide effective performance test strategies.

Claims

1. A method for creating workload model to test performance of an application in a data processing network, wherein the data processing network comprises a plurality of data processing units, the method comprising:

receiving, at one of the plurality of data processing units, an activity log with a plurality of entities from the data processing network, wherein each entity of the activity log comprises a session identification number;

creating a transaction log of all transactions conducted using the activity log, wherein a user defines at least one transaction with respect to a particular session identification number, and wherein the at least one transaction is a business transaction flow, the business transaction flow comprising a user defined logical sequence of destination addresses;

determining a plurality of transaction metrics using the transaction log, wherein the plurality of transaction metrics comprises an average transaction response time, a transaction throughput for a session identification number, and a total transaction throughput for all session identification numbers, wherein the determination of the transaction throughput for a session identification number and the total transaction throughput for all the session identification numbers comprises;

a. transaction throughput for a session identification number ‘

X’

=number of transaction for ‘

X’

/(Σ

(elapsed time of each transaction))

b. total transaction throughput for all the session identification numbers=(Σ

(transaction throughput for all session identification numbers)/(total number of session identification numbers), wherein the total transaction throughput is the number of transactions executed in a certain time period; and

creating a workload model to test the performance of the application in the data processing network based on the plurality of transaction metrics and provide performance test strategies.

2. The method as recited in claim 1, wherein the activity log is maintained at a plurality of data processing units in the data processing network.

3. The method as recited in claim 1, wherein each entity of the activity log further comprises a date, a logged time, a time taken and a destination address.

4. The method as recited in claim 1, wherein the transaction log is created from the activity log using a parsing algorithm, the parsing algorithm comprising:

segregating the plurality of entities for the particular session identification number from the activity log;

determining if a first destination address from the segregated entities corresponding to the particular session identification number matches with a first destination address of at least one of the transactions defined by the user;

further determining, only after the first destination addresses are found matching, if a second destination address from the segregated entities corresponding to the particular session identification number matches with a second destination address of at least one of the transactions as identified in the previous step; and

continuing in the same manner till all the destination addresses from the segregated entities corresponding to the particular session identification number match with all the destination addresses of at most one transaction defined by the user, ensuring the first and the last destination addresses in the plurality of entities corresponding to any session identification number are not repeated at all while the destination addresses in between the first and the last destination addresses are repeated as many times although in the same sequence.

5. The method as recited in claim 1, wherein said determining the plurality of transaction metrics comprises determining an elapsed time, defined as the difference, in terms of unit time, between the time when the request for the first destination address in a transaction is received by the data processing unit and the response for the last destination address in a transaction is sent from a data processing unit.

6. The method as recited in claim 1, wherein said determining the plurality of transaction metrics comprises determining transaction level think time, defined as the summation of all destination address level think times, a destination address level think time defined as the difference, in terms of unit time, between the time when the response for a previous destination address is sent by the data processing unit and the request for the current destination address is received by the data processing unit.

7. The method as recited in claim 1, wherein said determining the plurality of transaction metrics comprises determining response time of data processing unit, defined as the summation of the entries under “

time taken”

field in the activity log for each destination address in a transaction.

8. The method as recited in claim 3, wherein the destination address in the activity log is a Uniform Resource Locator (URL) of a webpage from the application in the data processing network.

9. A system capable of creating a workload model in order to test performance of an application in a data processing network, wherein the data processing network comprises a plurality of data processing units, the system comprising:

a receiving module to receive an activity log with a plurality of entities from the data processing network, wherein each entity of the activity log comprises a session identification number;

a processor configured to;

create a transaction log of all transactions conducted using the activity log, wherein a user defines at least one transaction with respect to a particular session identification number, and wherein the at least one transaction is a business transaction flow, the business transaction flow comprising a user defined logical sequence of destination addresses; and

determine a plurality of transaction metrics using the transaction log, wherein the plurality of transaction metrics comprises an average transaction response time, a transaction throughput for a session identification number, and a total transaction throughput for all session identification numbers, wherein the determination of the transaction throughput for a session identification number and the total transaction throughput for all the session identification numbers comprises;

a. transaction throughput for a session identification number ‘

X’

=number of transaction for ‘

X’

/(Σ

(elapsed time of each transaction))

b. total transaction throughput for all the session identification numbers=(Σ

(transaction throughput for all session identification numbers)/(total number of session identification numbers), wherein the total transaction throughput is the number of transactions executed in a certain time period; and

a workload model generating module to create a workload model to test the performance of the application in the data processing network based on the plurality of transaction metrics and provide performance test strategies.

10. The system as recited in claim 9, wherein receiving module automatically retrieves the activity log from the memory of the plurality of data processing units in the data processing network.

11. The system as recited in claim 9, wherein each entity of the activity log further comprises a date, a logged time, a time taken and a destination address.

12. The system as recited in claim 9, wherein the processor creates the transaction log from the activity log using a parsing algorithm, the processor is further configured to:

segregate the plurality of entities for the particular session identification number from the activity log;

determine if a first destination address from the segregated entities corresponding to the particular session identification number matches with a first destination address of at least one of the transactions defined by the user;

further determine, only after the first destination addresses are found matching, if a second destination address from the segregated entities corresponding to the particular session identification number matches with a second destination address of at least one of the transactions defined by the user as identified in the previous step; and

continue in the same manner till all the destination addresses from the segregated entities corresponding to the particular session identification number match with all the destination addresses of at most one transaction defined by the user, ensuring the first and the last destination addresses in the plurality of entities corresponding to any session identification number are not repeated at all while the destination addresses in between the first and the last destination addresses are repeated as many times although in the same sequence.

13. The system as recited in claim 9, wherein the processor determines the plurality of transaction metrics, the processor is configured to determine an elapsed time, defined as the difference, in terms of unit time, between the time when the request for the first destination address in a transaction is received by the data processing unit and the response for the last destination address in a transaction is sent from a data processing unit.

14. The system as recited in claim 9, wherein the processor determines the plurality of transaction metrics, the processor is configured to determine transaction level think time, defined as the summation of all destination address level think times, a destination address level think time defined as the difference, in terms of unit time, between the time when the response for a previous destination address is sent by the data processing unit and the request for the current destination address is received by the data processing unit.

15. The system as recited in claim 9, wherein the processor determines the plurality of transaction metrics, the processor is configured to determine response time of data processing unit, defined as the summation of the entries under “

time taken”

field in the activity log for each destination address in a transaction.

16. The system as recited in claim 11, wherein the destination address in the activity log is a Uniform Resource Locator (URL) of a webpage from the application in the data processing network.

17. A computer storage device tangibly embodying a plurality of instructions adapted for creating a workload model in order to test performance of an application in a data processing network, wherein the data processing network comprises a plurality of data processing units, the computer storage device comprising:

a computer readable medium storing;

program code adapted for receiving, at a computing device, an activity log with a plurality of entities from the data processing network, wherein each entity of the activity log comprises a date, a logged time, a time taken, a session identification number, and a destination address;

program code adapted for creating a transaction log of all transactions conducted using the activity log, wherein a user defines at least one transaction with respect to a particular session identification number, and wherein the at least one transaction is a business transaction flow, the business transaction flow comprising a user defined logical sequence of destination addresses;

program code adapted for determining a plurality of transaction metrics using the transaction log, wherein the plurality of transaction metrics comprises an average transaction response time, a transaction throughput for a session identification number and a total transaction throughput for all session identification numbers wherein the determination of the transaction throughput for a session identification number and the total transaction throughput for all the session identification numbers comprises;

a. transaction throughput for a session identification number ‘

X’

=number of transaction for ‘

X’

/(Σ

(elapsed time of each transaction))

b. total transaction throughput for all the session identification numbers=(Σ

(transaction throughput for all session identification numbers)/(total number of session identification numbers), wherein the total transaction throughput is the number of transactions executed in a certain time period; and

program code adapted for creating a workload model to test the performance of the application in the data processing network based on the plurality of transaction metrics and provide performance test strategies.