Experience sharing of automatic test for students with monthly salary of 28K

When we formulate the implementation strategy of automated testing , First of all, we should consider the possible risks .

Insufficient time for automated testing

Expectations for automated testing are too high

Lack of experience in automated test implementation

Automated testing tools are updated too often

Automated testing tools do not help software testing itself

After we have prepared for the risks of automated test implementation , You can start thinking ：

At what stage do you need to start automated testing ？

When to start automated testing , Every company is different . Some companies are when the test cases are manually executed and the test cases are no longer modified , Then develop the corresponding automated test script ;

And some companies are developing test cases at the same time , Develop scripts . If the test case designer in the team is an engineer with rich experience in test case design , The test cases he developed are efficient , Few changes in the future , You can consider developing test cases at the same time , Synchronous development of automated test scripts .

If the test case designer in the team is a person who is not experienced in test case design or whose test case quality is not efficient , The developed test cases need to be changed frequently in the later stage , Then consider waiting until the test case itself is stable , Then start script development .

How about the human input of automated testing ？

Most companies develop automated test scripts by dedicated personnel , A few large companies develop automated test scripts for all . Both methods have their own advantages and disadvantages ： Specially assigned person for script development , The advantage is that the professional skills of developing scripts can be continuously strengthened , Development efficiency is greatly improved ; The disadvantage is that the test cases of the development module are not well understood , It is possible to develop automated test scripts just “ translate ” The test case , Find out bug It's less likely .

And some big companies , Because the overall quality of employees is high , Usually have a certain development ability , The manual tester of each module develops the automatic test script by himself . although , Manual testers are less proficient in script development than specialized script developers , However, manual testers are the ones who know the true meaning of test cases best , So the test script he developed is not just “ translate ”, It may be the test case “ sublimation ”, Its test script found bug The probability will be greater .

How to execute the test script more efficiently ？

（1）N Test environments execute test scripts synchronously and in parallel , The total time cost of automated test script execution can be reduced to 1/N.

（2） A dedicated automated test execution engineer executes batch automated test scripts . Before the automated test script fails to run 3 The major factors are ：

Test environment issues

Script error

The measured target appears bug

Due to the accumulation of specialized automated test execution engineers' experience in analyzing a large number of failed scripts , It is usually very efficient to locate the cause of script failure , Improve the efficiency of automated test script execution .

（3） An independent automated test environment for script execution teams . As mentioned earlier , The test environment problem is the reason why the test script fails . The test environment affects the two killers of test script execution ： One is that the test environment was destroyed by the previous failed script and was not restored ; The other is that the testing environment is destroyed by colleagues on other projects .

For the first case , We can add enough system recovery code to the code structure of the test script to solve ; For the second case , Only rely on the policy support of the company's leaders , Are you willing to free up enough testing environment for the automation test execution team .

（4） Add rich positioning information of script failure to the test script . Once the automated test script fails , We can only rely on the information printed by the script itself to locate , The speed of locating problems depends on the experience of script executors , It depends more on whether there is rich script printing information in the script .

（5） Use automated test baseline software version . When a large number of test scripts fail , After eliminating the problem of test environment , Directly run these failed test scripts in the baseline software version .

If it runs in the baseline version and passes , It proves that the reason for the script failure is that the product is new bug Caused by the , Instead of reading the source code of these failed test scripts one by one to analyze the reasons of the script itself .

Now I invite you to join our software testing learning exchange group ：【746506216】, remarks “ The group of ”, We can discuss communication software testing together , Learn software testing together 、 Interview and other aspects of software testing , There will also be free live classes , Gain more testing skills , Let's advance together Python automated testing / Test Development , On the road to high pay .

Friends who like software testing , If my blog helps you 、 If you like my blog content , please “ give the thumbs-up ” “ Comment on ” “ Collection ” One Key triple connection ！