Data warehouse test analysis method based on blood relationship

Abstract

The invention relates to a data warehouse test analysis method based on a blood relationship. The method comprises the steps of 1, analyzing a tested data demand to obtain a target table set; establishing a corresponding relation table between tables and hierarchies, namely identifying a configuration table; performing identification and blood relationship analysis on the target table set according to the identification configuration table to obtain a source table on which the target table set depends and a table set of each layer in the data warehouse; 2, establishing a use case generation rule, and matching the source table in the step A and each layer of table set in the data warehouse with the use case generation rule to obtain a test use case containing the source table, each layer oftable set in the data warehouse and a target table. According to the method, the efficiency is improved, and the problem that vulnerability is likely to occur when a traditional manual analysis source table is used is eliminated; meanwhile, the possible unreliability of the test analysis method of taking the interior of the data warehouse as a black box can be solved, and the problems of test data preparation omission, test execution failure, test result verification failure and the like caused by analysis omission are prevented.

Description

technical field [0001] The invention relates to the field of software testing, in particular to a blood relationship-based data warehouse testing and analysis method. Background technique [0002] A data warehouse is a subject-oriented, integrated, relatively stable data collection that reflects historical changes and is used to support management decisions. Build a data warehouse to store and monitor real-time data and historical data; the data warehouse usually adopts a hierarchical structure design, and each upstream level has a one-to-one or many-to-one relationship with its adjacent downstream levels. [0003] With the rapid development of big data technology in recent years, the traditional software testing theory lacks special testing and analysis methods for data warehouses. The testing of data warehouses is essentially to simulate business scenarios, extract, convert and process the generated source table data, and obtain target table data. Data, and then test and ...

AI Technical Summary

Problems solved by technology

[0004] Traditionally, there are two schemes for analyzing the data flow between the source table and each level of the data warehouse; scheme 1: analyze the target table according to the data requirements, manually write the test cases based on the target table, and at the same time, according to the requirements description and the experience of the testers, Manually determine the source table involved, while the data transfer at each level between the source table and the target table is regarded as a black box. Solution 1 analyzes the source table based on the data requirements and the experience of the testers, which is easy to analyze and omit, resulting in test rework; solution Two: Analyze the target table according to the data requirements, manually write test cases based on the target table, do not care about which source tables are involved, and do not care about the data flow from the source table to the middle layers of the target table, describe the simulated business scenario according to the requirements, and then Perform a full amount of data extraction, conversion and processing, and finally test and verify the target table data according to the test cases. Solution 2 for full amount of data extraction has extremely high requirements for the test environment and consumes a lot of work. It is very difficult to implement when the data warehouse is large. Difficult; the above two solutions only test, verify and analyze the results of the target table, and cannot analyze all levels inside the data warehouse, and it is easy to miss hidden problems

Images