Design Scheme of High Concurrent and High Usable Cache based on TB Level Data Capacity

Abstract

With the continuous development of the Internet in recent years, the Internet business has also changed greatly. In the past, when accessing the database, it usually takes the way of direct access, in order to resist the reading and writing traffic, often use a master and data fragmentation methods. However, the current traffic is more and more, and the amount of data is also accumulating. The data capacity of many e-commerce websites is getting bigger and bigger, even exceeding the TB level. If we still use the previous way, all the traffic using the database to bear, its stability will be greatly reduced, and it will also bring higher cost and lower efficiency, is extremely undesirable. Therefore, there is an urgent need to find technologies and architectures related to high concurrency and high availability. For the above reasons, this topic first understands and analyzes the previous application platform cache and Redis database cluster cache, and then builds a multi-level cache strategy based on Njinx local cache, which will be described from requirements analysis and architecture design.