To meet the stringent service requirements of 6G applications such as immersive cloud eXtended Reality (XR), holographic communication, and digital twin, there is no doubt that an increasing number of servers will be deployed on the network edge. Then, the techniques, edge computing, edge caching, and edge intelligence will be more widely utilized for intelligent local data storage and processing generated by 6G applications, while innovative access network architecture based on the cloud-edge servers, such as the Open-Radio Access Network (O-RAN) will be adopted to improve the flexibility and openness for new service deployment and frequent network changes. On the other hand, new attack surfaces and vectors targeting local infrastructure and users will emerge along with the deployment of novel network architecture and techniques. Massive researchers have studied the potential security and privacy threats on the 6G network edge as well as the countermeasures. The three techniques, edge computing, edge caching, and edge intelligence have become a double-edged sword that can not only be synchronously utilized to develop defense countermeasures, but also become the targets of many new security and privacy threats. In this article, we provide a comprehensive survey of articles on the three techniques-related security threats and countermeasures on the 6G network edge. We explain how security and privacy can be destroyed by attacking one of the three technologies and how the three services support each other to realize efficient and achievable security protection. Moreover, the researches on the benefits and limitations of Federated Learning (FL) and blockchain for decentralized edge network systems in terms of security and privacy are also investigated. Additionally, we also analyze the existing challenges and future directions towards 6G.