Hi, I would request feedback from experienced users regarding if the approach that I am using for my "BI" project is on the right track or otherwise. Here's how I am doing the various steps: 1) In the OLTP, I have defined views which "make available" the data that needs to be brought into my data mart 2) On each run, my DTS's (bringing in the data from the OLTP into my data mart), drop and create the tables and then populate them using the views mentioned above... The reason for dropping and creating the table is becuase in the early days (when I started this learning process) I was having problems can recall the details but these were related to my time dimension and also trying to figure out what data to append and what not to append.... 3) In my data mart, I have created views on which the various dimensions and fact tables are based 4) I am doing a FULL process of dimension and cubes rather than incremental. I am doing this hoping that this is the right approach and that this will ensure that everything goes right. 5) I have used MOLAP all across. Should I consider other options if performance is the key criteria? 6) All my dimesions are shared dimesions. 7) I have set the performance to 100% I would appreciate some comments from theexperts against the various stages that I have listed above and how I am going about each. Will appreciate if you can share your recommendations regarding considering alternatives based on your experience. Many TIA.