Design And Implementation Of Cache System Based On Memcached

With the high develop of network technology, high-performance data cache system has become the research focus of e-commerce platform and telecommunications service providers. It is also an important problem and challenge that the other high-speed growth of business enterprises has to face, its low latency and efficient and high concurrency so that the overall design of the system performance in all aspects have put forward a high requirements. With the high develop of network technology, memory and variety of hardware facilities are becoming cheaper, high-performance data cache system is gotten more and more attention.The paper first introduces the cache technology, including the principle of caching technology, classification and common replacement algorithm. Then introduce a high-performance cache system in detail, including the network topology of cache system, communication protocols, and some distributed cache systems. Then, based on the analysis of memcached, we design a high-performance cache system which based on the memcached. In the designing of system, we should consider the network topology, cache replacement and hot spot in the system. Memcached cache system is divided into three levels, namely the network layer, system core layer and storage layer. Network layer is responsible for external client connections and connections management, the main function is to interpret the command and handle high concurrent connections. System core layer is responsible for the entire core functionality of the distributed system; the main function includes data storage, load balancing, and managing the life cycle of multi-copy of the hot spots. In the system core layer, in order to prevent the local peak access, presents an effective management strategy of multi-replica. Through this strategy can not only prevent a single server becoming the bottleneck of the system, and can make full use of the other servers. The storage layer is exactly a memcached node, responsible for the storage of data.In the scenarios of high-performance caching system, we test function and structure of the system on the rational design of the test, which shows that the designed system has good practicability.