Sorry about reposting this here as a new thread, but maybe the post was so deep in the thread that nobody pay attention to it. Here it goes again. I'm trying to speed up the cube processing time, by means of partitions. Let's suppose I have defined the partitions in such a way that I'm getting processed just the more recent records. So I process just the partition with the last added records. The problem is with the virtual cubes that rely on my just recent processed cube. How can I avoid processing all the virtual cube? there is no partitions in the virtual cube..... All the time I gain processing just the partition with new data is lost when I'm forced to do full process (or refresh data) in the virtual cube. Thanx.