Partition Level Statistics gathering

Question

Hi,
This is with regards to GRANULARITY option in DBMS_STATS.GATHER_TABLE_STATS procedure.
If I give 'PARTITION' as GRANULARITY would it gather statistics for all the partitions on a table or can I gather statistics only for a particular partition.
For example: If I have table T1 which is partitioned by month i.e JAN08, FEB08, ....., DEC09, JAN09, ... and If I apply DML only on say DEC08, can I gather statistics only for DEC08 partition ?
Please advise.
Thanks in Advance.
Regards
Pat

Answer 1

I do not think it is just because of misspelled word PARTITION:
SQL> CREATE TABLE "Patza"
  2      ( prod_id        NUMBER(6)
  3      , cust_id        NUMBER
  4      , time_id        DATE
  5      , channel_id     CHAR(1)
  6      , promo_id       NUMBER(6)
  7      , quantity_sold  NUMBER(3)
  8      , amount_sold         NUMBER(10,2)
  9      )
10  PARTITION BY RANGE (time_id)
11    (PARTITION "Sales_Q1_1998" VALUES LESS THAN (TO_DATE('01-APR-1998','DD-MON-YYYY')),
12     PARTITION "Sales_Q2_1998" VALUES LESS THAN (TO_DATE('01-JUL-1998','DD-MON-YYYY')),
13     PARTITION "Sales_Q3_1998" VALUES LESS THAN (TO_DATE('01-OCT-1998','DD-MON-YYYY')),
14     PARTITION "Sales_Q4_1998" VALUES LESS THAN (TO_DATE('01-JAN-1999','DD-MON-YYYY')),
15     PARTITION "Sales_Q1_1999" VALUES LESS THAN (TO_DATE('01-APR-1999','DD-MON-YYYY')),
16     PARTITION "Sales_Q2_1999" VALUES LESS THAN (TO_DATE('01-JUL-1999','DD-MON-YYYY')),
17     PARTITION "Sales_Q3_1999" VALUES LESS THAN (TO_DATE('01-OCT-1999','DD-MON-YYYY')),
18     PARTITION "Sales_Q4_1999" VALUES LESS THAN (TO_DATE('01-JAN-2000','DD-MON-YYYY')),
19     PARTITION "Sales_Q1_2000" VALUES LESS THAN (TO_DATE('01-APR-2000','DD-MON-YYYY')),
20     PARTITION "Sales_Q2_2000" VALUES LESS THAN (TO_DATE('01-JUL-2000','DD-MON-YYYY')),
21     PARTITION "Sales_Q3_2000" VALUES LESS THAN (TO_DATE('01-OCT-2000','DD-MON-YYYY')),
22     PARTITION "Sales_Q4_2000" VALUES LESS THAN (MAXVALUE))
23  ;
Table created.
SQL> BEGIN
  2  DBMS_STATS.GATHER_TABLE_STATS
  3  (
  4  OWNNAME => USER
  5  , TABNAME => 'Patza'
  6  , ESTIMATE_PERCENT => 25
  7  , METHOD_OPT => 'FOR ALL COLUMNS SIZE AUTO'
  8  , GRANULARITY => 'PATRITION'
  9  , PARTNAME => 'Sales_Q3_2000'
10  , CASCADE => TRUE
11  );
12  END;
13  /
BEGIN
ERROR at line 1:
ORA-20000: Unable to analyze TABLE "SCOTT"."PATZA" SALES_Q3_2000, insufficient
privileges or does not exist
ORA-06512: at "SYS.DBMS_STATS", line 13046
ORA-06512: at "SYS.DBMS_STATS", line 13076
ORA-06512: at line 2
SQL> BEGIN
  2  DBMS_STATS.GATHER_TABLE_STATS
  3  (
  4  OWNNAME => USER
  5  , TABNAME => '"Patza"'
  6  , ESTIMATE_PERCENT => 25
  7  , METHOD_OPT => 'FOR ALL COLUMNS SIZE AUTO'
  8  , GRANULARITY => 'PATRITION'
  9  , PARTNAME => '"Sales_Q3_2000"'
10  , CASCADE => TRUE
11  );
12  END;
13  /
BEGIN
ERROR at line 1:
ORA-20001: Illegal granularity PATRITION: must be AUTO | ALL | GLOBAL |
PARTITION | SUBPARTITION | GLOBAL AND PARTITION
ORA-06512: at "SYS.DBMS_STATS", line 13056
ORA-06512: at "SYS.DBMS_STATS", line 13076
ORA-06512: at line 2
SQL> BEGIN
  2  DBMS_STATS.GATHER_TABLE_STATS
  3  (
  4  OWNNAME => USER
  5  , TABNAME => 'Patza'
  6  , ESTIMATE_PERCENT => 25
  7  , METHOD_OPT => 'FOR ALL COLUMNS SIZE AUTO'
  8  , GRANULARITY => 'PARTITION'
  9  , PARTNAME => 'Sales_Q3_2000'
10  , CASCADE => TRUE
11  );
12  END;
13  /
BEGIN
ERROR at line 1:
ORA-20000: Unable to analyze TABLE "SCOTT"."PATZA" SALES_Q3_2000, insufficient
privileges or does not exist
ORA-06512: at "SYS.DBMS_STATS", line 13046
ORA-06512: at "SYS.DBMS_STATS", line 13076
ORA-06512: at line 2
SQL>
SQL> BEGIN
  2  DBMS_STATS.GATHER_TABLE_STATS
  3  (
  4  OWNNAME => USER
  5  , TABNAME => '"Patza"'
  6  , ESTIMATE_PERCENT => 25
  7  , METHOD_OPT => 'FOR ALL COLUMNS SIZE AUTO'
  8  , GRANULARITY => 'PARTITION'
  9  , PARTNAME => '"Sales_Q3_2000"'
10  , CASCADE => TRUE
11  );
12  END;
13  /
PL/SQL procedure successfully completed.
SQL> As you can see, table/partition check is done first. So OP either does not have privs on the table or, as I mentioned, table/partition names are sensitive.
SY.