Software testing and quality assurance are complex and systematic projects, and the improvement of technology, people, processes, standards and other links is indispensable. This requires test practitioners to possess very comprehensive skills, especially as they move to higher-level positions, the more complex and challenging the issues they need to care about and solve.
The reason why the test masters become masters is not only because of their strong technical strength, but more importantly, they have experienced wind and rain, and are good at summing up the blood and tear experience of various projects. They can quickly improve their comprehensive capabilities in actual combat.
This article will share the test improvement Tips that a test veteran has continuously summarized from his work for your test students' reference! Also welcome to leave a message in the comment area to share your test improvement opinions!
Test Improvement Tips
1. Test a product or project, not only to test the function and performance of the product, but also to understand the various situations (the whole call volume, the peak value from the overall perspective (increasing to the product perspective)? Customer usage, which customers ? Online cluster deployment, etc.), know why a certain module was added?
2. If you find a problem, you need to ask 5 why? (Refer to 5 Why) Go deeper and deeper, recognize the root cause, and implement solutions.
For example, you find that your website is broken. Obviously, your first reaction is to restore it. But once the crisis has passed, a post-mortem investigation should be conducted to ask why:
Why did the website crash? Because the CPU usage on all front-end servers has reached 100%.
Why is the CPU usage reaching 100%? Because a new piece of code contains an infinite loop.
Why is there such a code? Because an employee made a mistake.
Why did he make such an error? Because he didn't write unit test code for this function.
Why didn't he write unit test code? Because he is a new employee and has not received proper training in test-driven development.
Make corrections in these five places:
restore the website.
Remove the error code.
Help an employee who made a mistake to understand why his code did not work as expected.
Train the employee to learn knowledge about test-driven development.
Incorporate test-driven development into the introduction guide for new engineers.
3. The fundamental purpose of everything you do is to solve the pain points, whether it is automated tests, tools or processes. Do n’t forget your original intention, do n’t forget why you left because you left too long.
4. The reason and background of the development modification point must be clearly understood (including the specific code).
5. Pay attention to the project's data model (table structure) and technical implementation (design idea) for project risk assessment.
6. Take over the test work of a project. As an owner, in addition to clarifying the requirements, you also need to know: when the project is going to go online, what function points need to go online, and what percentage? Is it consistent with the plan? Then, evaluate whether it is reasonable. If it is urgent, you need to consider the corresponding test solution.
7. Design test cases, in addition to coverage of functional scenarios, we must also consider the characteristics of the project itself. (For example, concurrency at the gateway layer, connection timeout, stability, etc., such as idempotency of Redis commands, high availability of Kafka, consistency, etc.)
8. The test report specifies the cause and solution of the bug.
9. Improve the stability of automated testing (set timeout, error retry, etc.).
10. Everyone has their own main task. For the test tool (platform), its practicality and user experience are very important. Because only when these tools can really help users (and reduce maintenance costs), will they "live" for a long time. (Avoid uselessness)
11. If something goes wrong, don't be afraid. The focus is on quickly locating bugs and solving them quickly. Emphasize the importance of logs and alarms.
12. Communicate with the development: If you are busy, you can wait, but give a clear time.
13. Technology is the foundation, but in the final analysis the ability to solve problems is the core force. (To learn to break through the fixed model and thinking limitation, analyze and locate the cause of the problem from multiple angles)
14. Review of online faults. (Understand the background, the cause of occurrence, and how to avoid it)
15. Data capability: tool or platform generation-> promotion and application (formation of data)-> analysis and optimization (output capability)
16. Many technologies, tools or products are universal, and it is the bright spot that requires breakthroughs for specific scenarios of the company.
(The article is from Hogwarts School of Testing)