While the web applications are reaching new horizons, applications are been accessed by hundreds and thousands of user at a time. Moreover it is highly mandate to ensure that the application is consistent in its performance in terms of its availability, accessibility and timely response from the server. Performance Testing can be divided into two sub process: Load Testing †It involves number of users access the application simultaneously to create load on the application and confirm existence of any system bottleneck if any. Stress Testing †It involves number of users accessing the application in batches to put stress on the application to the extent that system breaks down after a certain limit of stress. While screening an application for its performance certain parameters are crucial to measure performance of an application during peak loads: Response Time Web server Connectivity Database Connectivity SMTP server connectivity Data transfer rate from client to server and vice versa Number of pages and sessions performed along with number of hits The performance of the application should be almost same for one user and thousands of users accessing it at the same time varying only by few milliseconds. Performance Testing procedures are carried out using Performance Testing Tools which uses Virtual Users to replace the real time users accessing the application where the limit of number of Virtual Users is specified while creating load scenarios. There are number of tools available ranging from Opensource like OpneSta and paid ones like WAPT or LoadRunner. The following procedures are implemented for a seamless and successful performance testing of any application: Record User Scenarios Parameterize Test Scripts Group User Scenarios Create Load Scenarios Simulate and Execute Load Analyze Test Results Reports The application used for performance testing should be either production application or near to production having no or minor functional defects, it should be supported by web servers and load balancer depending upon number of virtual users needed to execute the test. Duration for any Performance test cycle can only be defined after understanding overall scope of the application, complexity of the application, number of testing scenarios to be considered, number of virtual users etc.
Please Rate this Article 5 out of 54 out of 53 out of 52 out of 51 out of 5
Not yet Rated
