The OLTP system can not afford the demand of decision support due to the drastic commerce competition, so the appearance of OLAP technology is importanct. The key technical of OLAP, the classical resolvent of data warehouse and the methods of data preprocessing are studied. A frame of parallel data extraction interface for commercial OLAP is proposed. Data preprocessing technology (data cleaning, data integration and transform, data reduction) connected with the frame is practically studied and applied. The eficiency and the practicality of the frame are illustrated by practical application.