Re: How is possible that avg_data_blocks_per_key = 0 even with fresh statistics ?

To throw away every doubts I regathered statistics now, but results are the same:

SQL> begin dbms_stats.gather_table_stats(user, 'T_DIPARTIMENTI_BUDGET', cascade => true, method_opt => 'for all columns size 1');
  2 end;
  3 /

PL/SQL procedure successfully completed

SQL> select count(*)
  2 from t_dipartimenti_budget;

COUNT(*)

SQL> select s.num_rows, s.blocks, s.last_analyzed   2 from user_tables s
  3 where s.table_name = 'T_DIPARTIMENTI_BUDGET';

NUM_ROWS BLOCKS LAST_ANALYZED
-------- ---------- -------------

      73 13 06/12/2007 18.11.11

SQL> select s.avg_data_blocks_per_key, s.blevel, s.leaf_blocks, s.last_analyzed, s.distinct_keys
  2 from user_indexes s
  3 where s.index_name = 'PK_T_DIPARTIMENTI_BUDGET';

AVG_DATA_BLOCKS_PER_KEY BLEVEL LEAF_BLOCKS LAST_ANALYZED DISTINCT_KEYS

----------------------- ------ ----------- ------------- -------------
                      0      0           0 06/12/2007 18 0

On Dec 6, 5:53 pm, hpuxrac <johnbhur..._at_sbcglobal.net> wrote:
> On Dec 6, 10:42 am, nicola.far..._at_info-line.it wrote:
>
> > DBMS_STATS.GATHER_SCHEMA_STATS(user, METHOD_OPT => 'for all columns
> > size 1', CASCADE => True);
>
> > scheduled each day early morning
>
> Dollar gets you a doughnut that the table was empty at the time you
> analyzed it.
Received on Thu Dec 06 2007 - 11:12:11 CST