Fighting with bugs in software is a hard work. Hundreds and thousands of testers trying to detect defects in software before running it on commercial use, otherwise the cost of errors will increase a thousand times. Reduce the cost of providing quality of products possible through the use of automated systems for troubleshooting.
Since most of the defects detected at the stage of product testing that determines to save money is to automate this stage of implementation. Company Mercury made interrogation of 1000 customers and found that approximately 80% of them do not use automation tools to test, preferring to do it manually. Of the remaining share the vast majority – 80% of companies – use only the simplest test automation tools during the performance of individual projects. In 14% of companies it is deployed special products testing and a standard infrastructure for this. Another 5% of companies have implemented projects and formed the testing centers of competence, aggregating best practices and exchange of experiences between teams and projects. And only 1% of customers implemented a system of total quality control and run the centralized testing services using a single life cycle for all projects.
With the Help of What Can We Test?
The leading product in the market test automation is currently HP Quality Center. In the Quality Center test automation products include: Control Panel Dashboard, a means of organizing and conducting testing, Test Director, means of functional testing and testing business processes.
Test Director is the core of Quality Center. It contains five modules, tightly integrated with each other and ensures the continuity of the testing process – Service Test Manager, Requirement Manager, Test Plan, Test Lab and Defect Manager. With it help you can define the structure of testing, to establish a process for interaction between different actors and to ensure the issuance of documentation.
In addition, Quality Center includes functional testing tools: Quick Test Professional – actively developing a new product aimed at a classic set of applications that is most often used in business practices and supports the latest technologies (SOA, etc.).
WinRunner works with old applications with an interface GUI. This product is gradually fades into the background and will not develop any more.
How to Automate Software Testing?
Starting point for automated testing of software products is the understanding and formalization of the functional requirements, which makes the customer to the service or software product. These requirements govern the typical actions of the operator. When forming a list of demands, naturally, first of all take into account all critical steps, both in terms of business process and in terms of consequences.
These actions will be checked more thoroughly, both independently and as part of the entire business process. To do this, create a work scenario, considering different aspects of the business process step by step, scroll through all the options that can be foreseen, up to his connections with other business processes.
For each of the claims it is creating a test that checks the correctness of system responses. All tests linked the testing procedures that can be both manual and automatic.
For example during manually testing the operator receives a message that he needs to open a certain window to perform a certain sequence of actions, and as a result he should see the expected result. At the end of the procedure, he asks the question – does he saw this result or not, which is fixed. In contrast, when run automatically test all the steps are performed by computer emulating user actions.