Oracle 11g changes to dbms_stats

Oracle Database Tips by Donald Burleson

In Oracle 11g we now see 11g extended optimizer statistics, an alternative to dynamic_sampling for estimating result set sizes.

Arup Nanda has a great article on extended statistics with dbms_stats, specialty histogram analysis using function-based columnar data:

Next, re-gather statistics on the table and collect the extended statistics on the expression upper(cust_name).

begin

  dbms_stats.gather_table_stats (

     ownname    => 'ARUP',

     tabname    => 'CUSTOMERS',

     method_opt => 'for all columns size skewonly for columns (upper(cust_name))'

);

end;

Alternatively you can define the column group as part of the gather statistics command. You do that by placing these columns in the method_opt parameter of the gather_table_stats procedure in dbms_stats as shown below:

begin
   dbms_stats.gather_table_stats (
      ownname         => 'ARUP',
      tabname         => 'BOOKINGS',
      estimate_percent=> 100,
      method_opt  => 'FOR ALL COLUMNS SIZE SKEWONLY FOR COLUMNS(HOTEL_ID,RATE_CATEGORY)',
       cascade         => true

The new 10g default value for dbms_stats method_opt is 'FOR ALL COLUMNS SIZE AUTO'.

Using "for all columns size auto" directs Oracle to decide automatically which will get histograms. Oracle examine the skew of column values and also examines the "workload" associated with the columns (querying v$sql for current SQL). However, the overhead of creating and managing zillions of histograms can easily outweigh the performance benefits of having the histograms.

It's a perpetual trade-off between sub-optimal plans and unneeded overhead.

An exploration of "intelligent histogram creation", is needed, a method that uses AWR to correlate the SQL to the objects, avoiding histograms that are never used, and develop a method to only create histograms that "make a difference".

��