Distributing data across partitions

Creating a clustered index on a partitioned table redistributes the table’s data evenly over the partitions. Adaptive Server determines the index key ranges for each partition so that it can distribute the rows equally in the partition. Each partition is assigned at least one exclusive device if the number of devices in the segment is equal to or greater than the number of partitions.

If you create the clustered index on an empty partitioned table, Adaptive Server prints a warning advising you to re-create the clustered index after loading data into the table, as all the data will be inserted into the first partition until you re-create the clustered index.

If you partition a table that already has a clustered index, all pages in the table are assigned to the first partition. The alter table...partition command succeeds and prints a warning. You must drop and recreate the index to redistribute the data.