Adaptive Server keeps statistics about the distribution of pages within a partitioned table and uses these statistics when considering whether to use a parallel scan in query processing. When you partition a table, Adaptive Server stores information about the data pages in each partition in the control page.
The statistics for a partitioned table may become inaccurate if any of the following occurs:
The table is unpartitioned and then immediately repartitioned.
A large number of rows are deleted.
A large number of rows are updated, and the updates are not in-place updates.
A large number of rows are bulk copied into some of the partitions using parallel bulk copy.
Inserts are frequently rolled back.
If you suspect that query plans may be less than optimal due to incorrect statistics, run the update partition statistics command to update the information in the control page.
The update partition statistics command updates information about the number of pages in each partition for a partitioned table.
The update all statistics command also updates partition statistics.
Re-creating the clustered index or running reorg rebuild automatically redistributes the data within partitions and updates the partition statistics. dbcc checktable, dbcc checkdb, and dbcc checkalloc also update partition statistics as they perform checks.
