Multidimensional interval querying method and system thereof

Abstract

The invention relates to a multidimensional interval querying method and a system thereof. The method comprises the following steps of: 1. organizing a transcript for copying data to be a plurality of complementary cluster index tables which are mutually compensated and verified, wherein each complementary cluster index table builds a sequence table which takes the length of a column value, an original line primary key and a column value as a primary key by means of each index column, completely stores the data of the other columns in the original line, and is used for continuously scanning when querying; and 2. converting a query string into a query plane tree, and performing query optimization to complete the query. The invention can simultaneously meet the requirements of high performance, low storage cost and high reliability.

Description

technical field [0001] The invention relates to the field of computer processing, in particular to a multi-dimensional interval query system and method. Background technique [0002] Resource discovery requires queries with multiple conditions, most of which can be converted into multidimensional range queries. In fact, multi-dimensional interval query is a basic requirement of network applications. A simple example is the application of storing Internet image information. The designer may need to query the top 100 images with the highest number of clicks within a period of time. This query involves multiple attributes such as time and clicks (the columns of the data table, also called dimension, dimension) interval. [0003] As the amount of network application data continues to increase, it is difficult for existing methods to meet the requirements of high performance, low storage overhead and high reliability at the same time. By classifying the existing data models, i...

AI Technical Summary

Problems solved by technology

This method will generate random read operations when retrieving data tables, and the speed of random read operations in distributed sequential tables is an order of magnitude lower than that of sequential scans (scan), so under this index mechanism, the multi-dimensional range query The speed is slow, and the response time is too long to meet the needs of current network applications for real-time retrieval of massive data

If you do not build a secondary index, but build a clustered index for the index column to improve query performance, the storage overhead will increase exponentially due to the copy of the underlying file system

Images