Common Problems of test automation

Everyone has understood and recognized the role of the testing tool. However, many software companies that have introduced the automated testing tool have not been able to make the test automation play its due role, the main reasons are as follows:

1.Incorrect ideas or unrealistic expectations

There is no correct idea of automated software testing, or it is too urgent, or the idea that automated testing can replace manual testing, or that automated testing can discover a large number of new defects, or you do not pay enough attention to it, rather than invest a large amount of money in the initial stage. In most cases, there are too many optimistic and high expectations for software testing automation. People expect that this test automation solution can solve all the problems currently encountered. At the same time, software vendors for testing tools will naturally emphasize the advantages, advantages, or successes of their tools. They may not mention the persistent efforts and difficulties they need to achieve such success. As a result, the initial expectation is not fulfilled.

 

2Lack of testing talents with good quality and experience

Some software companies are willing to spend 100,000 yuan to buy testing tool software, but lack of testing talents with good quality and experience. Software testing automation does not simply use testing tools, but also requires good testing procedures and comprehensive test cases to be compiled in concert with the script, this requires testers to be familiar with not only the product features and application fields, but also the testing process, as well as the testing and programming technologies..

 

3. Problems with the testing tool affect the testing quality.
Generally, the automatic test script is not tested on a large scale. Therefore, the quality of the Automatic Test script is often dependent on the experience and attitude of Ta engineers, if the automatic test tool does not provide a mechanism to ensure the script quality, the correctness of the test results will be directly affected.

The test case tested by the automated testing tool does not need to be manually tested. It can be effectively combined with the manual test, and the results of the automated test are also reflected in the final test report, is a more correct approach.

 

4No effective and adequate training

Personnel and Training complement each other. Without good, effective, and adequate training, testers lack depth and breadth of understanding of testing tools, resulting in inefficient use, the application results are not ideal. This kind of training is a long-term process, and it can be achieved through one or two lectures. In addition, in the actual use of the test tool, the user of the test tool may still have such a problem, which also requires a dedicated person to solve. Otherwise, will seriously affect the enthusiasm of the test tool.

5.Blindly introduce testing tools without considering the company's actual situation
It is clear that different testing tools are designed for different testing purposes and have their own characteristics and applicability. Therefore, not any good testing tool can adapt to the needs of different companies. A company has spent a lot of money introducing testing tools with good wishes. Half a year later, the testing tool became a decoration. The reason is that the actual situation of the company is not taken into account, and testing tools are unrealistically expected to change the current situation of the company, resulting in failure.

For example, most software companies in China develop projects for end users-Software of the engineering nature, rather than product development. The project development cycle is short, and different user requirements are different. In addition, the demand and user interface changes greatly throughout the development process. In this case, it is not suitable for introducing black box testing software, because the basic principle of the black box testing software is to record/play back (although through modification, a structured test script is formed), for constantly changing needs and interfaces, the workload of modifying and recording Scripts may greatly exceed the workload of Testing implementation. Using testing tools not only reduces the workload, but also increases the workload of testers. In this case, you can consider introducing white box testing tools to improve code quality.

 

6.No good environment for using test tools

To establish a good application environment for testing tools, the test process and management mechanism must be changed. Only in this way can the testing tool truly play its role. For example, for automated testing based on GUI recording/playback, changes to the product interface have a great impact on the normal operation of the script. Furthermore, white-box testing tools are generally used in the unit testing phase, while unit testing is done by developers in most companies. If there is no process to regulate developers' behaviors, when the project progress is relatively high, developers may consciously avoid using test tools to avoid problems. Therefore, it is necessary to clarify the use of test tools in the development and testing processes. For example, in the documents submitted for each milestone of the project, the reports generated by some test tools must be included, for example, the test coverage report generated by the devpartner tool during integration testing and the code quality report generated by logiscope.

7. Other technical and organizational issues

The test script required for software testing automation requires a large amount of maintenance, and the code changes of the software product must comply with certain rules, so as to ensure good repeatability of the test script, that is to say, test automation and software products cannot be separated.

Second, third-party vendors that provide software testing tools lack sufficient understanding of customers' applications, making it difficult to provide strong technical support and specific problem-solving capabilities. That is to say, there will be more or less problems with the interoperability between software testing tools and tested objects-software products or systems. In addition, with the ever-changing technical environment, all these applications for testing automation will be promoted and deepened, it has a huge impact.

There is also the illusion of security. If the software testing tool does not find any defects in the tested software, it does not indicate that there is no problem in the software, the test tool may be incomplete or the expected test results may be set incorrectly.