Automation testing: Devil time
Let me tell you a scenario where automation is useful. Now consider testing websites that have 3 modules, login, registration, and veranda. Suppose if there is a bug in the registration stream to enter to reach the veranda, after a bug fix, QA must register, enter, and check the Home page content. To run this flow, let's assume the QA takes about 5 minutes. But if the QA uses automation, this process will not be time-consuming> 1 minute.
I agree to write an automatic test case it will take an early time. Let's consider writing a test case that will take around 15 minutes at first. So to run this complete flow for the first time it will take about 16 minutes. But for testing the next website, QA only requires a single minute order. So to test the website 5 times after each bug fix or regression testing, the total number of times needed by automation is 21 minutes. But in the case of 25 minutes, manual testing is only needed if the QA pays attention to the system.
Now you have to think why it should be 4 minutes to save the problem. As a QA in the real world we must carry out the same test cases many times and sometimes their exhaustion, causing a decrease in testing efficiency.
From here alone, I can conclude 2 main reasons why automation can be useful. The time saved is the money saved and automation saves time as listed in the example above. The second is the efficiency of the testing system. Automation ensures efficiency will remain high at any time.
Automation testing: reluctance
Yes, agree there are many positive sides to automation but why still the company adopts automation testing as a standard let's look into it.
1 - Proficiency is needed to write automation test scripts
2 - debugging script is the main problem.
3 - Maintenance of expensive tests if the playback method. Although small changes occur in the graphical user interface (GUI), the test script must be re-recorded or replaced with a new test script.
4 - Maintenance of test data files is difficult if the test script tests more screens. This is the reason, having a strong data log service provider is the need for a mandate. Make sure you get support from one before continuing with the testing process and type of analysis and other log solving logs.
In general, everything in the cycle, both the development of Android applications, testing or even bugs has its own life cycle. Next, after talking about automation testing for the last few words, let's talk about the life cycle considered when automation testing is seen. Meanwhile, keeping a strong solution for Android logging as well as mandatory.
1. Automation Feasibility Study - When it comes to automation, ROI is an important factor, this feasibility study provides stakeholders and all business partners who automate this project will be useful or not.
2. Develop evidence of concepts - now feasibility studies do provide a high level of understanding of project automation, proof of concept reports include many statistics and information that states supporting automation which in turn provides confidence to stakeholders.
3. Evaluation of Automation Tools - The market today contains various tools for automation. Some tools are free to use and some are paid. Similarly, there are tools that require skilled personal and some do not require skills to appear. So this stage discusses the strategy of which tools are most suitable for the project.
No comments:
Post a Comment