Subscribe by Email


Monday, December 17, 2012

Give an overview of the process of system performance validation with IBM Rational Performance Tester?


IBM brought the rational performance tester in to picture as an effort to automate the rigorous process of performance testing. Automated performance testing was required to make it possible to deliver high quality software to the end users.

- The Rational Performance Tester is a convenient means for accelerating the software system’s performance without impacting its quality. 
- Software performance bottlenecks can be identified very easily with the help of IBM rational performance tester. 
- Not only the presence of bottlenecks but also their causes can be identified.  
It is important to make predictions regarding the behavior and performance of the software systems and applications under extreme load conditions. 
Making wrong predictions can get the whole organization, its revenue as well as the customer’s at risk. 
This can prove to be hell devastating in today’s software engineering world where the competition is already soaring high. 

- Hence, a proper testing solution is required that would not only help you validate the performance of your product but also optimize and verify it under a number of load conditions. 
- IBM rational performance tester gives you more than just testing. 
- Using it, the scalability factor of many of the applications can be calculated. 
The rational performance tester supports code free programming i.e., the programming knowledge is not much required. 

In this article we provide an overview of the system performance validation using the rational performance tester. 

- Performance tests are contained within performance test projects.
- Instead of creating a new project, a test can be simply recorded which in turn will create a project immediately. 
- After a test has been recorded, it can be edited so as to include data pools, data correlation and verification points. 
- Data pools are required for obtaining variable data. 
- At verification points, it is confirmed whether the test is running as expected or not. 
- Lastly the data correlation ensures that appropriate data is being returned for the request. 
- Furthermore, protocol specific elements can also be added to a test.
- The items to which you make modifications appear in italics whenever a test is edited. 
- After saving the test it changes in to regular type. 
- A workload can be emulated by adding user groups and other elements to a new schedule.
- After being done with all the additions to the schedule, it can be run. 
- You can either run it locally or with a default launch configuration. 
- The playback might be slowed down because of the delays in HTTP requests. - The performance results of the application can be evaluated after being generated dynamically during the execution process. 
- After execution, the results can be regenerated for analyzation and viewing purpose. 
- The coding problems which are known to cause disturbances in the performance can be found and fixed using the problem analysis tools. 
- The data regarding the response time breakdown can be obtained using an application that has been deployed in a production environment. 
- Another means for collecting this data is from a distributed application implemented in a testing environment. 
- Resource monitoring data including available network, disk usage, network throughput, processor usage and so on can also be collected and imported. 
Resource monitoring data is important since one can have a comprehensive view of the system or application and this is helpful in the determination of problems.
- All the collected and imported data can be analyzed and later can be used in the location of the exact cause of the problem that is hampering the system performance.



No comments:

Facebook activity