Load test monitoring system

• US 7,133,805 B1
• Filed: 07/07/2004
• Issued: 11/07/2006
• Est. Priority Date: 07/07/2004
• Status: Expired due to Fees

• Current Assignee: Sprint Communications Company LP (Deutsche Telekom AG)
• Original Assignee: Sprint Communications Company LP (Deutsche Telekom AG)
• Inventors: Willis, Gary D. Jr., Micklavzina, Charles L., Dankenbring, David W.
• Primary Examiner(s): Hoff, Marc S.
• Assistant Examiner(s): Huynh, Phuong
• Application Number:
  US10/886,250
• Time in Patent Office: 853 Days
• Field of Search: 702/186, 702182-185, 702/187, 702/188, 714/47
• US Class Current: 702/186
• CPC Class Codes: G06F 11/3414 Workload generation, e.g. scripts, playback
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/80 Database-specific techniques
• CPC Class Codes: G06F 2201/87 Monitoring of transactions

Abstract

A system for load test monitoring a computer system is provided. The system includes a transaction generator to provide transactions to servers to load test the servers. The system includes one or more agents provided on the servers, the agents operable to obtain server statistics from servers during load testing. The system also includes a display including a plurality of gauges. Each gauge depicts transactions statistics based on the transactions generated by the transaction generator for each server and further depicts server statistics for each server as obtained by the agents. Based on the displayed transaction and server statistics the transaction generator is adjustable to optimize the load test of the servers.

Claims

1. A system to monitor a load test of a computer system, comprising:

a plurality of agents, each operable on at least one of a plurality of application servers to report server statistics of each of the application servers;

a monitor operable to receive at least some of the server statistics from the at least some of the plurality of agents;

a load generation tool operable to send a plurality of transactions to load test one or more of the plurality of application servers, the load generation tool operable to generate a display for each of the plurality of application servers, the display containing at least one transaction statistic of one of the application servers;

a screen analyzer operable to capture as an image a portion of at least one of the displays containing the transaction statistic for one of the application servers and to convert the image into a text number; and

a dashboard component operable to receive the server statistics from the monitor and to receive the at least one transaction statistics from the screen analyzer as a text number, the dashboard component further operable to generate a dashboard display including the server statistics and the transaction statistics;

wherein the plurality of agents are further defined as plurality of performance agents and a plurality of communications agents, each operable on at least one of the plurality of application servers the performance agents operable to report a performance statistic of each of the application servers, the communication agents operable to report a communication statistic of each of the application servers;

wherein the monitor is further defined as a plurality of monitors each operable to receive performance statistics from at least one of the performance agents and further receive communication statistics from at least one of the communication agents, the monitor operable to receive performance statistics and communication statistics from the performance and communication agents; and

wherein the received communication statistics comprise the message queue depth of each of the application servers.

2. The system of claim 1 wherein the performance agent periodically reports the at least one performance statistic of the application server, the communication agent periodically reports the at least one communication statistic of the application server, the screen analyzer periodically reports the at least one transaction statistic, and the dashboard component periodically generates the dashboard display of the performance statistics, the communication statistics, and the transaction statistics.

3. The system of claim 2 wherein the dashboard component generates the dashboard display in joint photographic experts group format and transmits the dashboard display to a web server.

4. The system of claim 3 wherein the dashboard display is saved along with a time stamp each time the dashboard display is generated, and wherein a sequence of dashboard displays comprising each of the saved dashboard displays is available for rapid playback.

5. The system of claim 1 wherein the load generation tool is the Mercury LoadRunner tool and the screen analyzer is the SnagIt screen capture tool.

6. The system of claim 1 wherein the performance statistics is selected from a group consisting of a central processor unit utilization statistic, a memory utilization statistic, and a disk utilization statistic.

7. The system of claim 1 wherein the transaction statistics are selected from the group consisting of a transactions per hour, a transactions per second, a path failure, and a response time.

8. The system of claim 1 further including a results reporter operable to analyze logs generated by the application server and logs generated by the load generation tool to produce one or more reports on the load test.

9. The system of claim 8 wherein one of the reports includes a graphical display representing the range of transaction response times as a percentage of a transaction response time service level agreement.

10. The system of claim 9 wherein the graphical display further provides an indication of a response time of the last test transaction and an indication of a average response time of the final five test transactions.

11. A method for monitoring a load test of a computer system, comprising:

generating a plurality of transactions and sending the plurality of transactions to a plurality of application servers from a load test tool;

designating a first application server of the plurality of servers;

collecting at least one communication statistic from the first application server as a text number;

collecting at least one performance statistic from the first application server as a text number;

displaying, using the load test tool, transaction statistics related to the transactions generated to test at least one of the plurality of application servers;

capturing a view of the display of the transaction statistics using a capture tool;

analyzing the view of the display with the capture tool to render the view of the display as a text to identify a text number related to the transaction statistics;

providing the text numbers from the communication statistic, the performance statistic, and the transaction statistics to a dashboard component, the dashboard component operable to generate a gauge using the text numbers;

displaying the gauge associated with the first application server illustrating the at least one communication statistic, the at least one performance statistic, and the at least one transaction statistic based on the captured text numbers; and

providing the display the gauge to a web site accessible to monitor the load test.

12. The method of claim 11 wherein the collecting the at least one communication statistic, the collecting the at least one performance statistic, the collecting the at least one transaction statistic, the providing the text numbers to a dashboard component, the generating the model of the gauge, and the providing the display based on the model of the gauge to a web site is repeated on a periodic basis.

13. The method of claim 12 further comprising:

storing the display of the gauge along with a timestamp associated the display; and

displaying a sequence of the stored displays of the gauge in rapid playback.

14. The method of claim 11 further including:

observing the display of the gauge; and

adjusting the load generation tool to optimize load testing of the computer system.

15. The method of claim 14 wherein adjusting the load generation tool includes causing the load generation tool to generate more transactions per unit time.

16. The method of claim 11 wherein the at least one communication statistic is a queue depth statistic.

17. The method of claim 11 wherein the at least one performance statistic selected from a group consisting of a central processing unit utilization statistic, a memory utilization statistic, and a disk utilization statistic.

18. The method of claim 11 wherein the at least one transaction statistic selected from a group consisting of a transactions per hour statistic, a transactions per second statistic, a path failure statistic, and a response time statistic.

19. The method of claim 11 further including generating, after the completion of load testing the computer system, one or more reports based on logs generated by the first application server and based on logs generated by the load test tool.

20. A system for load testing and monitoring a computer system, comprising:

a transaction generator to provide transactions to servers to load test the servers;

one or more agents provided on the servers, the agents operable to obtain server statistics from servers during load testing;

a dashboard component operable to receive transactions statistics as text numbers based on the transactions generated by the transaction generator for each server and server statistics as text numbers from the one or more agents; and

a display generated by the dashboard component including a plurality of gauges, each gauge depicting transactions statistics based on the transactions generated by the transaction generator for each server and further depicting server statistics for each server as obtained by the agents, and wherein based on the displayed transaction and server statistics the transaction generator is adjustable to optimize the load test of the servers.

21. The system of claim 20 wherein the server statistics are further defined as server performance and communication statistics.