摘要
目前,图数据广泛应用于金融科技、智能问答、智慧城市等多个领域。图数据库作为图数据的存储介质,技术发展迅猛。随着数据量的爆发式增加,图中顶点和边的数量随之增加,所需要的存储和计算能力远超出单台机器所能提供的能力,因此分布式图数据库成为图数据库发展的必然方向。JanusGraph作为一种开源图数据库,具备可拓展节点数量的分布式存储能力,支持海量点边图数据的存储,在分布式环境下的图查询和图遍历都有着良好的性能。虽然JanusGraph支持数据的分布式存储,但是其服务本身Janus-Server并未支持高可用。文章提出一种基于JanusGraph的高可用服务架构,补足JanusGraph的这一缺陷,提升JanusGraph的高可用性。
Currently,graph data is widely used in many fields such as financial technology,intelligent question answering,and smart city.Graph databases,as the storage medium for graph data,have seen rapid technological development.With the explosive increase in the amount of data,the number of vertices and edges in the graph has also become huge,and the storage and computation capabilities required by these graphs are far beyond the capabilities provided by a single machine,and distributed graph databases have become an inevitable direction for the development of graph databases.JanusGraph,as an open source graph database,has the ability to expand the number of nodes for distributed storage,supports the storage of massive pointedge graph data,and has good performance in graph query and graph traversal in distributed environments.Distributed environment of the graph query and graph traversal have good performance.While JanusGraph supports distributed storage of data,its service itself,Janus-Server,does not support high availability.In this paper,we propose a high-availability service architecture based on JanusGraph to complement this defect of JanusGraph and enhance the high availability of JanusGraph.
作者
林萍
陈锡锻
李霄
LIN Ping;CHEN Xiduan;LI Xiao(Zhejiang Industry&Trade Vocational College,Wenzhou Zhejiang 325000,China)
出处
《信息与电脑》
2023年第21期164-166,共3页
Information & Computer
基金
2022年浙江工贸职业技术学院校级课题“基于动态图神经网络的消费金融反欺诈算法研究”(项目编号:纵20220005)。
关键词
图数据库
分布式架构
高可用
graph database
distributed architecture
highly available