Performance testing Overview

Obstacles to Performance Testing implementation
-Performance tests are often executed only after they are executed.
-Out-of-the-box Phenomenon
-Some meaningless tests will be performed.
-Incomplete test environment
-Insufficient full-layer staff engagement
-Expensive tools

User perspective
• Response Time

Development Perspective
• Comprehensive Requirement Analysis
• System Architecture Rationality
• Code execution efficiency

System Perspective
• System capacity
• Concurrency
• Stability
• Scalability

The fault tolerance test includes two tests:

1. Enter abnormal data or perform abnormal operations to verify the system's protection. If the system has good fault tolerance, the system only provides prompts or internal digestion, without causing system errors or even crashes.
2. Disaster recovery testing. The software is forced to fail through various means, and then the user data stored in the system is verified to be lost, and the system and data can be quickly restored.

The concept of automatic recovery test shows that fault tolerance test is a type of adversarial test process. How to transfer and recover useful data when the test software fails. Failover is used to ensure that the test object can successfully transfer a key part of the running system or system to another device when a fault occurs, that is to say, the Standby System will "replace" the faulty system to avoid losing any data or transactions without affecting users' usage. To conduct a comprehensive test of failover, a good method is to describe all the objects of the test system with a system structure diagram to design test cases for all possible fault points in the diagram. The key object of spof in the system structure diagram is the design defect.