I have the compression Job running for a Infocube which holds millions of records. Due to system maintenence this BI_COMP*** job cancelled in the middle. This Infocube already compressed for year 2005 and the records in E fact table has the count 2.5 million. At present the compression ran for the year 2006 and cancelled because of maintenance. I could observe that the E fact table count increased to 3.2 million and the F fact table count remains same, meaning some 2006 records got upated in the E fact table and F fact table also holds all 2006 records. If i reran the compression again for 2006, it will double the entries for already updated records in E fact table. Is there anyway to resolve this problem without deleting the Infocube contents and reload the data.