A data management method and system, and data analysis device

Abstract

The invention discloses a method and a system for managing data, and a data analyzing device. The method includes: collecting structured query language (SQL) statements executed by an application program, and performing persistent storage; decomposing the SQL statements to obtain access objects of the SQL statements; recording accessed times of each access object in preset time; and extracting corresponding access objects as hot point data according to threshold value of the accessed times. According to the method and the system for managing the data, and the data analyzing device, the accessed times of the access objects of the SQL statements are obtained by analyzing the SQL statements executed by the application program, and accordingly the hot point data is obtained. Use status of logic layer objects (such as a chart and an index) is accurately reflected by finding the hot point data on a data application layer, and furthermore the hot point data can be accurately managed, system processing response time is decreased, system processing corresponding speed is improved, demand for high-end storage devices is reduced, and mass data management cost is reduced.

Description

technical field [0001] The invention relates to a data management technology, in particular to a data management method and system, and a data analysis device. Background technique [0002] With the rapid development of my country's telecommunications industry, especially the continuous growth of mobile communication users and broadband users, various telecom operators store and manage various massive data. Data life cycle management is an effective way to solve massive data management. [0003] Because the traditional data life cycle management only defines data life based on the time dimension, specifically, the data life cycle management is simply carried out in chronological order, such as bills, detailed bills, transaction records and other objects with time tags, Perform specific operations such as backup, archiving, and historical data migration in chronological order. [0004] The existing methods for implementing data lifecycle management according to the time dim...

AI Technical Summary

Problems solved by technology

[0005] 1. Poor applicability

[0006] The premise of managing data according to the time dimension is that the data has time tags such as bills and transaction disciplines. For those data that do not have time tags themselves, life cycle management cannot be performed according to traditional methods

[0007] There is currently no way to process these data without time tags. They can only be stored in online storage like firewood. The simple accumulation over time takes up valuable online high-end storage resources. At the same time, data management software, such as databases, etc. Day-to-day running creates a lot of stress

No one can tell whether these randomly stored data are in use and whether they can be backed up, archived and deleted.

[0008] 2. The system processing response time is long and slow

The data that the system needs to call may be stored on different levels of storage resources. The system calls data at a slow speed and takes a long time to respond

[0010] 3. A large number of high-end storage devices are required and the cost is high

[0011] In order to reduce the processing response time of the system, a large amount of data needs to be stored on the online high-end storage device, and the storage space of the online high-end storage device is relatively large, which greatly increases the cost of managing massive data

Images