Inverted Secondary Index Cluster Technique for Fast Query of Mass Quasi-Real-Time Data in Power Systems

For the problem that the growing scale of quasi-real-time data in power grids leads to the increasingly prominent obstacles of accessing it by using traditional relational databases, a secondary index technique is proposed by combining cluster framework with the inverted index structure. The processes of queries based on non-primary keys by the designed technique, first reverse locate the primary row key of the corresponding record with the help of inverted index cluster, and then complete the search in the NoSQL database cluster based on values of the primary row key. The designed technique integrates the advantages of inverted index and NoSQL database, and with the support of multi-host cluster, it can quickly complete the mass of quasi-real-time data query processing. To set the millions of engineering data as an example, the results show that the designed technique can reduce the time required to complete non-primary key query processing to hundred milliseconds.