Data storage method and device

Abstract

The embodiment of the invention discloses a data storage method and device. The method comprises the following steps: calculating hash values for identifying index fields of database partitions by adopting a preset algorithm; calculating hash values for identifying the IP (internet protocol) addresses of storage nodes by adopting a preset algorithm, wherein the hash values of the IP addresses of the storage nodes form a storage node ring; when data need to be stored, by taking the hash values of the index fields corresponding to the to be stored data as the starting points, searching the storage nodes from the storage node ring along the clockwise direction, and storing the to-be-stored data into the searched first storage node. According to the data storage method and device, the friendliness in the aspect of database partition extension and the balance of data storage can be ensured.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of data storage, and in particular, to a data storage method and device. Background technique [0002] With the widespread popularization of Internet technology, the e-commerce model has developed rapidly. With its advantages of low cost and high efficiency, the total amount of online business transactions has repeatedly hit new highs, but this has also brought about the problem of how to efficiently store and access massive data. The scale of the e-commerce database is constantly expanding, and it is under huge query pressure all the time. Therefore, many methods of database performance expansion have emerged. [0003] The most common method is to add database servers, copy multiple copies of the database and store them in different servers to achieve the purpose of dispersing query pressure, and use the principle that most database read operations are more frequent than write operatio...

AI Technical Summary

Problems solved by technology

[0006] Analyzing the existing horizontal partition technology, we can find that partitioning according to the hash value of the partition index field is relatively balanced, but it is not friendly to expansion. When the server needs to be expanded, all data must be repartitioned, which takes a lot of time; Segmentation according to the range of the index field, obviously, this method has good scalability, but it is easy to cause a large difference in the amount of data in each partition, which cannot achieve the effect of load balancing; create a database to store partition index fields to partitions The mapping relationship of the server is more flexible, but when the amount of image relationship data is large, the pressure on storage space is increased

Images