This paper, which is based on ETL, realized application of ETL technology in QHSE report system of PetroChina, and advanced solutions about large-scale data warehouse projects. It includes: data extract, transform, and load based on metadata, cleansing method of data. The detail is:1. Supply uniform data interface criterion for different business data source; It's obtained initial data files which included statistical reports and multi-dimension analyse from all fields of business system and work environment, manage data in reason to realize data conformity spanning multi-flat and multi-system;2. develop report system in C#.NET programming language, realize extracting of abnormity data files by XML technology;3. verify original data files, check out whether data accord with demand of data warehouse system and data validity, clean data and generate relative log files;4. load data files and data lines which passed validity verifying according to appoint demand and design load policy to elevate ETL efficiency.In the end of paper, I made summarize and prospect. Mainly,I summarized ETL method in data warehouse project development, and then point out key of research in the future, which is elevate efficiency of ETL further. |