Distributed systems are the pillars of the current computing ecosystem, which make modern computing more powerful, reliable, and flexible, covering several key fields from cloud computing and big data processing to the Internet of Things. However, due to the complexity of the system, some code defects are inevitably introduced during the code implementation of distributed systems, thus posing a huge threat to the availability, robustness, and security of the system. Therefore, the testing and defect detection work of distributed systems is very important. Dynamic testing technology conducts real-time analysis during the system operation to detect its defects and evaluate its behavior and functions, and is widely used in the defect detection of various system applications and has successfully found many code defects. A four-layer defect threat model of distributed systems is proposed in this study. Based on it, the testing requirements and main challenges of distributed systems are analyzed, and a general framework for dynamic testing of distributed systems is proposed. Then, typical dynamic testing tools for distributed systems are introduced from the perspective of detecting different types of system defects. Next, the study highlights critical techniques such as multidimensional test input generation, system-critical state awareness, and defect judgment criteria. Additionally, the paper reviews popular dynamic testing tools and evaluates their effectiveness in defect discovery and test coverage. The coverage and defect discovery capabilities of the current mainstream dynamic testing tools for distributed systems are evaluated. The findings show that multidimensional input generation significantly enhances testing efficiency. Finally, the study discusses emerging trends and future directions in dynamic testing of distributed systems, aiming to address their inherent challenges and improve testing outcomes.