ORACLE的统计信息在执行SQL的过程中扮演着非常重要的作用，而且ORACLE在表的各个层次都会有不同的统计信息，通过这些统计信息来描述表的，列的各种各样的统计信息。下面通过一个复合分区表来说明一些常见的和不常见的统计信息。
搭建测试用例
--创建表。 SQL> create table test partition by range(object_id) subpartition by hash(object_type) subpartitions 4 (partition p1 values less than(10000), partition p2 values less than(20000), partition p3 values less than(30000), partition p4 values less than(maxvalue)) as select * from dba_objects; --收集统计信息 sql> BEGIN dbms_stats.gather_table_stats( ownname => 'NC60', tabname => 'TEST', estimate_percent => 100, --百分之百采样 block_sample => FALSE, method_opt => 'FOR ALL COLUMNS SIZE 10', --收集直方图 granularity => 'ALL', --所有分区 cascade => TRUE --收集索引 ); END;
1,表级的统计信息
SQL> select table_name,num_rows,blocks,empty_blocks,avg_space 2 from user_tables 3 where table_name = 'TEST'; TABLE_NAME NUM_ROWS BLOCKS EMPTY_BLOCKS AVG_SPACE ------------------------------ ---------- ---------- ------------ ---------- TEST 87556 1328 0 0
2,表上列的统计信息
SQL> select table_name,column_name,num_distinct,density 2 from user_tab_columns 3 where table_name = 'TEST'; TABLE_NAME COLUMN_NAME NUM_DISTINCT DENSITY ------------------------------ ------------------------------ ------------ ---------- TEST OWNER 34 .023504483 TEST OBJECT_NAME 57495 .000020431 TEST SUBOBJECT_NAME 163 .007094837 TEST OBJECT_ID 87556 .000011421 TEST DATA_OBJECT_ID 22685 .000044786 TEST OBJECT_TYPE 45 .086650298 TEST CREATED 2057 .000847961 TEST LAST_DDL_TIME 1898 .000919514 TEST TIMESTAMP 2182 .000838526 TEST STATUS 1 5.7106E-06 TEST TEMPORARY 2 5.7106E-06 TABLE_NAME COLUMN_NAME NUM_DISTINCT DENSITY ------------------------------ ------------------------------ ------------ ---------- TEST GENERATED 2 5.7106E-06 TEST SECONDARY 2 5.7106E-06 TEST NAMESPACE 21 .116972867 TEST EDITION_NAME 0 0
3,表上列的直方图信息（OBJECT_ID列）
SQL> col TABLE_NAME format a20 SQL> col COLUMN_NAME format a40 SQL> select table_name,column_name,endpoint_number,endpoint_value 2 from user_tab_histograms 3 where table_name = 'TEST' 4 and column_name = 'OBJECT_ID'; TABLE_NAME COLUMN_NAME ENDPOINT_NUMBER ENDPOINT_VALUE -------------------- ---------------------------------------- --------------- -------------- TEST OBJECT_ID 0 2 TEST OBJECT_ID 1 8921 TEST OBJECT_ID 2 17998 TEST OBJECT_ID 3 26754 TEST OBJECT_ID 4 35510 TEST OBJECT_ID 5 44266 TEST OBJECT_ID 6 53025 TEST OBJECT_ID 7 62172 TEST OBJECT_ID 8 71290 TEST OBJECT_ID 9 82232 TEST OBJECT_ID 10 91577
4,分区的统计信息
SQL> select partition_name,num_rows,blocks,empty_blocks,avg_space 2 from user_tab_partitions 3 where table_name = 'TEST'; PARTITION_NAME NUM_ROWS BLOCKS EMPTY_BLOCKS AVG_SPACE ------------------------------ ---------- ---------- ------------ ---------- P1 9708 147 0 0 P2 9805 162 0 0 P3 10000 156 0 0 P4 58043 863 0 0
5,分区上列的统计信息
SQL> select column_name,num_distinct,density,num_nulls 2 from user_part_col_statistics 3 where table_name = 'TEST' 4 and partition_name = 'P1'; COLUMN_NAME NUM_DISTINCT DENSITY NUM_NULLS ---------------------------------------- ------------ ---------- ---------- OWNER 5 .000051504 0 OBJECT_NAME 7878 .000142267 0 SUBOBJECT_NAME 25 .00877193 9594 OBJECT_ID 9708 .000103008 0 DATA_OBJECT_ID 2074 .000560776 7597 OBJECT_TYPE 20 .080003882 0 CREATED 141 .011559584 0 LAST_DDL_TIME 213 .010730067 0 TIMESTAMP 182 .011371733 0 STATUS 1 .000051504 0 TEMPORARY 2 .000051504 0 COLUMN_NAME NUM_DISTINCT DENSITY NUM_NULLS ---------------------------------------- ------------ ---------- ---------- GENERATED 2 .000051504 0 SECONDARY 1 .000051504 0 NAMESPACE 8 .000051504 0 EDITION_NAME 0 0 9708
6,分区上列的直方图信息（OBJECT_ID列）
SQL> select column_name,bucket_number,endpoint_value 2 from user_part_histograms 3 where table_name = 'TEST' 4 and partition_name = 'P1' 5 and column_name = 'OBJECT_ID'; COLUMN_NAME BUCKET_NUMBER ENDPOINT_VALUE ---------------------------------------- ------------- -------------- OBJECT_ID 0 2 OBJECT_ID 1 1030 OBJECT_ID 2 2010 OBJECT_ID 3 2981 OBJECT_ID 4 3952 OBJECT_ID 5 4923 OBJECT_ID 6 5928 OBJECT_ID 7 6953 OBJECT_ID 8 7933 OBJECT_ID 9 8903 OBJECT_ID 10 9999 11 rows selected.
7,子分区的统计信息
SQL> select subpartition_name,num_rows,blocks,empty_blocks 2 from user_tab_subpartitions 3 where table_name = 'TEST' 4 and partition_name = 'P1'; SUBPARTITION_NAME NUM_ROWS BLOCKS EMPTY_BLOCKS ------------------------------ ---------- ---------- ------------ SYS_SUBP21 3314 48 0 SYS_SUBP22 3824 57 0 SYS_SUBP23 528 11 0 SYS_SUBP24 2042 31 0 SQL>
8,子分区上的列的统计信息
SQL> select column_name,num_distinct,density 2 from user_subpart_col_statistics 3 where table_name = 'TEST' 4 and subpartition_name = 'SYS_SUBP21'; COLUMN_NAME NUM_DISTINCT DENSITY ---------------------------------------- ------------ ---------- OWNER 3 .000150875 OBJECT_NAME 3314 .00030175 SUBOBJECT_NAME 1 .009615385 OBJECT_ID 3314 .00030175 DATA_OBJECT_ID 178 .005617978 OBJECT_TYPE 7 .000150875 CREATED 126 .017527186 LAST_DDL_TIME 148 .01653325 TIMESTAMP 134 .017483116 STATUS 1 .000150875 TEMPORARY 2 .000150875 COLUMN_NAME NUM_DISTINCT DENSITY ---------------------------------------- ------------ ---------- GENERATED 2 .000150875 SECONDARY 1 .000150875 NAMESPACE 5 .000150875 EDITION_NAME 0 0
9,子分区上的列的直方图信息
SQL> select column_name,bucket_number,endpoint_value 2 from user_subpart_histograms 3 where table_name = 'TEST' 4 and subpartition_name = 'SYS_SUBP21' 5 and column_name = 'OBJECT_ID'; COLUMN_NAME BUCKET_NUMBER ENDPOINT_VALUE ---------------------------------------- ------------- -------------- OBJECT_ID 0 100 OBJECT_ID 1 1764 OBJECT_ID 2 2429 OBJECT_ID 3 3088 OBJECT_ID 4 3752 OBJECT_ID 5 4309 OBJECT_ID 6 4640 OBJECT_ID 7 5828 OBJECT_ID 8 7278 OBJECT_ID 9 8912 OBJECT_ID 10 9998
我们对这个复合分区分析之后产生了上面这九种不同层次的统计信息。CBO想要得要一个高效的执行计划需要如此多的统计信息