For the problem that coal sale data has large volume but little information content, the paper proposed a design scheme of OLAP analyzing system of coal sale data based on Hadoop platform, introduced design ideas and structure of the system, and described specific process of implementation of deep and fast mining and intuitive display of data taking statistics of sales data as an example. The system uses Hadoop cloud platform for ETL data processing, creates Hive distributed data warehouse, and uses HQL language of Hive for OLAP statistical analysis, which achieves deep data mining, statistics and analysis of sales information from multi-level and multi-angle quickly and accurately, meanwhile reflects the result of data analysis intuitively form multi-angle.