Defining partition slice

As mentioned earlier, Analysis Services can intelligently decide which partitions to check for each query. For example, if we submit a query requesting reseller sales data for year 2005, SSAS should only have to check the Reseller Sales 2005 partition and not spend any time examining datafiles for 2006 or any other partitions. To exclude irrelevant partitions, Analysis Services checks the index files found within each partition folder; these files contain the range of internally assigned key values. You have already seen that we can partition measure groups by date; indeed, the majority of SSAS projects are partitioned by one of the levels found in the date dimension: day, month, quarter, or year. However, sometimes partitioning only at the date level is insufficient—you may have large volumes of intraday data that represents only a small portion of the total daily volume. In such case you can create partitions based on multiple attributes, for example, we could...