How does dbms_stats default granularity AUTO Work?

dbms_stats收集统计信息包的默认粒度为AUTO，对于AUTO没有非常明确的解释，一般认为它会收集分区的统计信息，但不包含子分区subpartition。对于这种说明我们加以核实：  

SQL> select * from v$version;

BANNER
----------------------------------------------------------------
Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bi
PL/SQL Release 10.2.0.5.0 - Production
CORE    10.2.0.5.0      Production
TNS for Linux: Version 10.2.0.5.0 - Production
NLSRTL Version 10.2.0.5.0 - Production

SQL>  select dbms_stats.get_param('cascade') from dual;
   select dbms_stats.get_param('degree') from dual;

DBMS_STATS.GET_PARAM('CASCADE')
--------------------------------------------------------------------------------
DBMS_STATS.AUTO_CASCADE

SQL> 
DBMS_STATS.GET_PARAM('DEGREE')
--------------------------------------------------------------------------------
NULL

SQL>    select dbms_stats.get_param('estimate_percent') from dual;

DBMS_STATS.GET_PARAM('ESTIMATE_PERCENT')
--------------------------------------------------------------------------------
DBMS_STATS.AUTO_SAMPLE_SIZE

SQL>    select dbms_stats.get_param('method_opt') from dual;

DBMS_STATS.GET_PARAM('METHOD_OPT')
--------------------------------------------------------------------------------
FOR ALL COLUMNS SIZE AUTO

SQL>    select dbms_stats.get_param('no_invalidate') from dual;

DBMS_STATS.GET_PARAM('NO_INVALIDATE')
--------------------------------------------------------------------------------
DBMS_STATS.AUTO_INVALIDATE

SQL>    select dbms_stats.get_param('granularity') from dual;

DBMS_STATS.GET_PARAM('GRANULARITY')
--------------------------------------------------------------------------------
AUTO

CREATE TABLE composite_rng_hash (
cust_id     NUMBER(10), 
cust_name   VARCHAR2(25), 
amount_sold NUMBER(10,2), 
time_id     DATE)
PARTITION BY RANGE(time_id) 
SUBPARTITION BY HASH(cust_id)
SUBPARTITION TEMPLATE(
SUBPARTITION sp1 TABLESPACE users,
SUBPARTITION sp2 TABLESPACE users,
SUBPARTITION sp3 TABLESPACE users,
SUBPARTITION sp4 TABLESPACE users) (
PARTITION sales_pre05
VALUES LESS THAN (TO_DATE('01/01/2005','DD/MM/YYYY')),
PARTITION sales_2005 
VALUES LESS THAN(TO_DATE('01/01/2006','DD/MM/YYYY')),
PARTITION sales_2006
VALUES LESS THAN(TO_DATE('01/01/2007','DD/MM/YYYY')),
PARTITION sales_2007
VALUES LESS THAN(TO_DATE('01/01/2008','DD/MM/YYYY')),
PARTITION sales_2008 
VALUES LESS THAN(TO_DATE('01/01/2009','DD/MM/YYYY')),
PARTITION sales_future
VALUES LESS THAN(MAXVALUE));

Table created.

SQL> select num_rows , blocks from dba_tables where  table_name='COMPOSITE_RNG_HASH';

  NUM_ROWS     BLOCKS
---------- ----------

SQL> select partition_name,num_rows , blocks from dba_tab_partitions where  table_name='COMPOSITE_RNG_HASH';

PARTITION_NAME                   NUM_ROWS     BLOCKS
------------------------------ ---------- ----------
SALES_PRE05
SALES_2005
SALES_2006
SALES_2007
SALES_2008
SALES_FUTURE

SQL> select partition_name,SUBPARTITION_NAME,num_rows , blocks from dba_tab_subpartitions where  table_name='COMPOSITE_RNG_HASH';

PARTITION_NAME                 SUBPARTITION_NAME                NUM_ROWS     BLOCKS
------------------------------ ------------------------------ ---------- ----------
SALES_PRE05                    SALES_PRE05_SP4
SALES_PRE05                    SALES_PRE05_SP3
SALES_PRE05                    SALES_PRE05_SP2
SALES_PRE05                    SALES_PRE05_SP1
SALES_2005                     SALES_2005_SP4
SALES_2005                     SALES_2005_SP3
SALES_2005                     SALES_2005_SP2
SALES_2005                     SALES_2005_SP1
SALES_2006                     SALES_2006_SP4
SALES_2006                     SALES_2006_SP3
SALES_2006                     SALES_2006_SP2
SALES_2006                     SALES_2006_SP1
SALES_2007                     SALES_2007_SP4
SALES_2007                     SALES_2007_SP3
SALES_2007                     SALES_2007_SP2
SALES_2007                     SALES_2007_SP1
SALES_2008                     SALES_2008_SP4
SALES_2008                     SALES_2008_SP3
SALES_2008                     SALES_2008_SP2
SALES_2008                     SALES_2008_SP1
SALES_FUTURE                   SALES_FUTURE_SP4
SALES_FUTURE                   SALES_FUTURE_SP3
SALES_FUTURE                   SALES_FUTURE_SP2
SALES_FUTURE                   SALES_FUTURE_SP1

24 rows selected.

SQL> exec dbms_stats.gather_table_stats('SYS','COMPOSITE_RNG_HASH');

PL/SQL procedure successfully completed.

SQL> select num_rows , blocks from dba_tables where  table_name='COMPOSITE_RNG_HASH';

  NUM_ROWS     BLOCKS
---------- ----------
         0          0

SQL> 
SQL> 
SQL>  select partition_name,num_rows , blocks from dba_tab_partitions where  table_name='COMPOSITE_RNG_HASH';

PARTITION_NAME                   NUM_ROWS     BLOCKS
------------------------------ ---------- ----------
SALES_PRE05                             0          0
SALES_2005                              0          0
SALES_2006                              0          0
SALES_2007                              0          0
SALES_2008                              0          0
SALES_FUTURE                            0          0

  1* select partition_name,SUBPARTITION_NAME,num_rows , blocks from dba_tab_subpartitions where  table_name='COMPOSITE_RNG_HASH'
SQL> /

PARTITION_NAME                 SUBPARTITION_NAME                NUM_ROWS     BLOCKS
------------------------------ ------------------------------ ---------- ----------
SALES_PRE05                    SALES_PRE05_SP4
SALES_PRE05                    SALES_PRE05_SP3
SALES_PRE05                    SALES_PRE05_SP2
SALES_PRE05                    SALES_PRE05_SP1
SALES_2005                     SALES_2005_SP4
SALES_2005                     SALES_2005_SP3
SALES_2005                     SALES_2005_SP2
SALES_2005                     SALES_2005_SP1
SALES_2006                     SALES_2006_SP4
SALES_2006                     SALES_2006_SP3
SALES_2006                     SALES_2006_SP2
SALES_2006                     SALES_2006_SP1
SALES_2007                     SALES_2007_SP4
SALES_2007                     SALES_2007_SP3
SALES_2007                     SALES_2007_SP2
SALES_2007                     SALES_2007_SP1
SALES_2008                     SALES_2008_SP4
SALES_2008                     SALES_2008_SP3
SALES_2008                     SALES_2008_SP2
SALES_2008                     SALES_2008_SP1
SALES_FUTURE                   SALES_FUTURE_SP4
SALES_FUTURE                   SALES_FUTURE_SP3
SALES_FUTURE                   SALES_FUTURE_SP2
SALES_FUTURE                   SALES_FUTURE_SP1

  以上证明了默认dbms_stats的GRANULARITY AUTO在常规情况下一般是收集分区统计信息， 而不收集子分区的统计信息。