Dynamic hybrid indexing method for non-relational database

Abstract

The invention relates to a dynamic hybrid indexing method for a non-relational database, which is characterized by comprising the following steps of establishing a non-primary key indexing structure of a key value pair non-relational database; defining the weight of the non-primary key field, and updating the weight value of the non-primary key field by taking the field as the frequency of the query condition and the historical weight of the field within one period; dynamically dividing index types; and carrying out hybrid indexing. According to the mixed indexing method of the key value pairnon-relational database provided by the invention, the conditional query efficiency of the non-primary key field can be effectively improved. The method overcomes the defect that the key value pair non-relational data does not support the non-primary key field and does not support the index query, and establishes the mixed index of the non-primary key field. In terms of big data storage, the function of quick query is achieved, and unnecessary storage expenditure is saved.

Description

technical field [0001] The invention relates to a method for establishing an index in big data retrieval, and belongs to the technical field of big data storage and retrieval. Background technique [0002] Distributed non-relational databases have high availability, high scalability, and high reliability, and are often used as storage support for big data. Distributed non-relational databases are divided into document type, key-value pair type, etc., and they have their own shortcomings. Some non-relational distributed databases of key-value pairs have dynamic scalability in fields and can flexibly store data of different structures, but they can only perform primary key index queries, and non-primary key queries can only scan and search. [0003] Many scholars have conducted research on this problem, and have also proposed some solutions to improve the retrieval speed. When storing some special data, some main fields can be encoded into the primary key to improve query ef...

AI Technical Summary

Problems solved by technology

[0009] The technical problem to be solved by the present invention is: the key-value pair non-relational database itself does not support non-primary key fields to build an index, although the non-primary key fields can be encoded and put into the primary key, so that the non-primary key fields can use index queries, but this The method is only suitable for data storage in a few special fields, and this method cannot support flexible index queries

Images