Dynamic data transferring method based on graph databases and graph database cluster

Abstract

The invention provides a dynamic data transferring method based on graph databases. The method includes the steps of dividing metadata into to-be-expanded data and non-to-be-expanded data according tothe relation of metadata of different types in the graph databases when the capacity of the graph databases serving as nodes in a cluster exceeds a preset threshold, and recording the to-be-expandeddata in to-be-expanded graph databases serving as newly-added nodes of the cluster. By introducing an arbitration module, the dynamic transferring of data between the multiple single-node graph databases is realized. When the capacity of data of a certain single-node graph database faces a bottleneck, database dynamic transferring is realized through the arbitration module, data on the nodes largein pressure is dynamically transferred to the nodes small in data pressure, and therefore the data balance among all the nodes is realized, the graph data storage capacity and the write performance linearity are increased, and meanwhile a corresponding graph database cluster is provided.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a graph database, and in particular to a graph database-based dynamic data migration method and a graph database cluster. Background technique [0002] Graph database refers to a type of NoSQL database, which is a non-relational database that uses graph theory to store relational information between entities. [0003] For example, the relationship between people in a social network, in the graph database, each person is represented as a point, and the relationship between people is represented by the edges between points and points. Traditional relational databases are not ideal for storing "relational" data, and often have the defects of complex and slow queries, while graph databases can perfectly make up for this defect. [0004] There are many existing graph databases, including but not limited to: Neo4j, Arangodb, Orientdb and other graph databases. [0005] Common...

AI Technical Summary

Problems solved by technology

[0006] However, most of the existing graph databases use the operation mode of synchronous writing and concurrent reading between single nodes.

Obviously, although this method can realize the linear expansion of the storage capacity and write performance of the graph database, it cannot improve the linear expansion of the read performance and has limitations.

Images