Cross-region-oriented multi-master-model distributed graph calculation method

Abstract

The invention provides a cross-region-oriented multi-master-model distributed graph calculation method, and relates to the technical field of graph calculation. The method comprises the following steps: forming a complete graph data set by taking all computing nodes in a plurality of data centers included in a cross-regional graph computing system as vertexes; determining a main vertex in the graph data set, and setting a copy vertex for the main vertex on the boundary of the graph data set; then establishing a graph calculation model, and performing state value updating on all non-boundary vertexes and boundary vertexes without copies in a graph data set; and establishing a multi-master calculation model on boundary vertexes based on a graph calculation model, and performing state updating on all master vertexes and duplicate vertexes thereof. Meanwhile, a message cache region is set for message sending of all vertexes in the graph calculation model and the multi-master calculation model so as to reduce the message sending amount. According to the method, the data privacy of each data center is ensured, and the global synchronization limitation is eliminated, so that each data center has higher autonomy.

Description

technical field [0001] The invention relates to the technical field of graph computing, in particular to a cross-regional multi-master model distributed graph computing method. Background technique [0002] Graph computing is an emerging big data mining technology, which is widely used in various industries, such as social network relationship mining, web link search, protein interaction function detection, road traffic network navigation analysis, etc. With the popularity of the global network and the continuous emergence of online applications, graph computing applications are deployed in data centers in various regions of the world and generate data in a geographically distributed manner. For example, social networks generate a large amount of text, voice, and video data every day. In order to reduce costs, these data are stored in data centers in the area near the data source; cellular networks collect data at geographically distributed base stations; in different region...

AI Technical Summary

Problems solved by technology

[0005] For the cross-regional graph computing system, it should solve the following three problems: 1. Wide area network (WAN) is used between data centers in various regions, and its bandwidth cost is expensive, and high costs may be incurred for transmitting large amounts of data

2. Data in most cross-regional data applications is sensitive

3. Autonomy of data management in each data center

Therefore, when the computing load of each data center is extremely skewed, adopting a synchronization strategy can only make the synchronization overhead more serious.

Secondly, mandatory collaborative processing between data centers does not conform to the idea of ​​cross-regional independent processing and regional autonomy

Images