11g 搜集直方图导致不走索引
2015-02-04 19:33
447 查看
SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_
WHERE messagein0_.msgID='ID:414d5120514d453333313020202020204c224a3c200c49cb';
1条记录
SQL_ID c1pvjs5wx132x, CHILD NUMBER 0
-------------------------------------
SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_
WHERE messagein0_.msgID=:1
PLAN HASH VALUE: 3093528534
-------------------------------------------------------------------------------
| ID | Operation | NAME | ROWS | Bytes | COST (%CPU)| TIME |
-------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | | | 71102 (100)| |
|* 1 | TABLE ACCESS FULL| MESSAGEIN | 910K| 47M| 71102 (1)| 00:14:14 |
-------------------------------------------------------------------------------
Predicate Information (IDENTIFIED BY operation ID):
---------------------------------------------------
1 - FILTER("MESSAGEIN0_"."MSGID"=:1)
SQL> set linesize 200
SQL> select a.column_name,
2 b.num_rows,
3 a.num_distinct Cardinality,
4 round(a.num_distinct / b.num_rows * 100, 2) selectivity,
5 a.histogram,
6 a.num_buckets
7 from dba_tab_col_statistics a, dba_tables b
8 where a.owner = b.owner
9 and a.table_name = b.table_name
10 and a.owner = 'MBFE'
11 and a.table_name = upper('messagein');
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 1650567 1650567 100 NONE 1
MSGID 1650567 1650567 100 FREQUENCY 2
STATUS 1650567 1 0 NONE 1
PROCCNT 1650567 2 0 NONE 1
ACCEPTTIME 1650567 1271808 77.05 HEIGHT BALANCED 254
MSGBODY 1650567 0 0 NONE 0
QUEUE 1650567 4 0 FREQUENCY 4
7 rows selected.
SELECT /*+ index (messagein0_ msgID_idx1)*/ messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_
WHERE messagein0_.msgID='ID:414d5120514d453333313020202020204c224a3c200c49cb';
SQL> select count(distinct(MSGID)) from mbfe.messagein;
COUNT(DISTINCT(MSGID))
----------------------
1721321
SQL> select count(*) from mbfe.messagein;
COUNT(*)
----------
1721321
BEGIN
DBMS_STATS.GATHER_TABLE_STATS(ownname => 'MBFE',
tabname => 'MESSAGEIN',
estimate_percent => 100,
method_opt => 'for columns MSGID size skewonly',
no_invalidate => FALSE,
degree => 16,
cascade => TRUE);
END;
只对end_dt搜集, repeat表示以前收集过直方图，现在收集统计信息的时候就收集直方图，如果以前没收集过直方图，现在收集统计信息的时候就不收集。
SQL> set linesize 200
SQL> select a.column_name,
2 b.num_rows,
3 a.num_distinct Cardinality,
4 round(a.num_distinct / b.num_rows * 100, 2) selectivity,
5 a.histogram,
6 a.num_buckets
7 from dba_tab_col_statistics a, dba_tables b
8 where a.owner = b.owner
9 and a.table_name = b.table_name
10 and a.owner = 'MBFE'
11 and a.table_name = upper('messagein');
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 1723263 1650567 95.78 NONE 1
MSGID 1723263 1723263 100 FREQUENCY 2
STATUS 1723263 1 0 NONE 1
PROCCNT 1723263 2 0 NONE 1
ACCEPTTIME 1723263 1271808 73.8 HEIGHT BALANCED 254
MSGBODY 1723263 0 0 NONE 0
QUEUE 1723263 4 0 FREQUENCY 4
7 rows selected.
SQL>
SQL> set linesize 200
SQL> set pagesize 200
SQL> explain plan for SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_
2 WHERE messagein0_.msgID='ID:414d5120514d453333313020202020204c224a3c200c49cb';
Explained.
SQL> select * from table(dbms_xplan.display());
PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 3093528534
-------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 928K| 48M| 73327 (1)| 00:14:40 |
|* 1 | TABLE ACCESS FULL| MESSAGEIN | 928K| 48M| 73327 (1)| 00:14:40 |
-------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - filter("MESSAGEIN0_"."MSGID"='ID:414d5120514d45333331302020202020
4c224a3c200c49cb')
14 rows selected.
SQL> BEGIN
2 DBMS_STATS.GATHER_TABLE_STATS(ownname => 'MBFE',
3 tabname => 'MESSAGEIN',
4 estimate_percent => 100,
5 method_opt => 'for columns MSGID size skewonly',
6 no_invalidate => FALSE,
7 degree => 16,
8 cascade => TRUE);
9 END;
10 /
PL/SQL procedure successfully completed.
SQL> select a.column_name,
2 b.num_rows,
3 a.num_distinct Cardinality,
4 round(a.num_distinct / b.num_rows * 100, 2) selectivity,
5 a.histogram,
6 a.num_buckets
7 from dba_tab_col_statistics a, dba_tables b
8 where a.owner = b.owner
9 and a.table_name = b.table_name
10 and a.owner = 'MBFE'
11 and a.table_name = upper('messagein');
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 1721323 1650567 95.89 NONE 1
MSGID 1721323 1721323 100 FREQUENCY 2
STATUS 1721323 1 0 NONE 1
PROCCNT 1721323 2 0 NONE 1
ACCEPTTIME 1721323 1271808 73.89 HEIGHT BALANCED 254
MSGBODY 1721323 0 0 NONE 0
QUEUE 1721323 4 0 FREQUENCY 4
SQL> select msgID from mbfe.messagein messagein0_ where rownum<10;
MSGID
----------------------------------------------------------------------------------------------------
ID:414d5120514d454d42464520202020204c24b44120263403
ID:414d5120514d453333313020202020204c224a3c200c1de7
ID:414d5120514d454d42464520202020204c24b44120263404
ID:414d5120514d453333313020202020204c224a3c200c49cb
ID:414d5120514d454d42464520202020204c24b44120263703
ID:414d5120514d453333313020202020204c224a3c200fa81d
ID:414d5120514d453333313020202020204c224a3c200e15d0
ID:414d5120514d453333313020202020204c224a3c200ff871
ID:414d5120514d453333313020202020204c224a3c200cd011
SQL> explain plan for SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_ where msgID='ID:414d5120514d454d42464520202020204c24b44120263403';
Explained.
SQL> select * from table(dbms_xplan.display());
PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 4204216909
--------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
--------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 55 | 4 (0)| 00:00:01 |
| 1 | TABLE ACCESS BY INDEX ROWID| MESSAGEIN | 1 | 55 | 4 (0)| 00:00:01 |
|* 2 | INDEX RANGE SCAN | INDEXMSGIN_1 | 1 | | 3 (0)| 00:00:01 |
--------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
2 - access("MSGID"='ID:414d5120514d454d42464520202020204c24b44120263403')
14 rows selected.
SQL> select a.column_name,
b.num_rows,
a.num_distinct Cardinality,
round(a.num_distinct / b.num_rows * 100, 2) selectivity,
a.histogram,
a.num_buckets
from dba_tab_col_statistics a, dba_tables b
where a.owner = b.owner
and a.table_name = b.table_name
and a.owner = 'MBFE'
and a.table_name = upper('messagein');
2 3 4 5 6 7 8 9 10 11
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 978619 978619 100 NONE 1
MSGID 978619 978619 100 NONE 1
STATUS 978619 1 0 NONE 1
PROCCNT 978619 2 0 NONE 1
ACCEPTTIME 978619 653888 66.82 NONE 1
MSGBODY 978619 0 0 NONE 0
QUEUE 978619 4 0 NONE 1
7 rows selected.
收集直方图后:
BEGIN
DBMS_STATS.GATHER_TABLE_STATS(ownname => 'MBFE',
tabname => 'MESSAGEIN',
estimate_percent => 100,
method_opt => 'for columns MSGID size skewonly',
no_invalidate => FALSE,
degree => 16,
cascade => TRUE);
END;
SQL> /
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 978619 978619 100 NONE 1
MSGID 978619 978619 100 FREQUENCY 2
STATUS 978619 1 0 NONE 1
PROCCNT 978619 2 0 NONE 1
ACCEPTTIME 978619 653888 66.82 NONE 1
MSGBODY 978619 0 0 NONE 0
QUEUE 978619 4 0 NONE 1
7 rows selected.
SQL> explain plan for SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_ where msgID='ID:414d5120514d454d42464520202020204c24b44120263403';
Explained.
SQL> select * from table(dbms_xplan.display());
PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 2678161237
-------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 454K| 23M| 43978 (1)| 00:08:48 |
|* 1 | TABLE ACCESS FULL| MESSAGEIN | 454K| 23M| 43978 (1)| 00:08:48 |
-------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - filter("MSGID"='ID:414d5120514d454d42464520202020204c24b441202634
03')
14 rows selected.
repeat表示以前收集过直方图，现在收集统计信息的时候就收集直方图，如果以前没收集过直方图，现在收集统计信息的时候就不收集。
skewonly 对所有的列搜集统计信息
干掉直方图信息:
BEGIN
DBMS_STATS.GATHER_TABLE_STATS(ownname => 'MBFE',
tabname => 'MESSAGEIN',
estimate_percent => 100,
method_opt => 'for columns MSGID size 1',
no_invalidate => FALSE,
degree => 16,
cascade => TRUE);
END;
SQL> select a.column_name,
b.num_rows,
a.num_distinct Cardinality,
round(a.num_distinct / b.num_rows * 100, 2) selectivity,
a.histogram,
a.num_buckets
from dba_tab_col_statistics a, dba_tables b
where a.owner = b.owner
and a.table_name = b.table_name
and a.owner = 'MBFE'
and a.table_name = upper('messagein'); 2 3 4 5 6 7 8 9 10 11
COLUMN_NAME NUM_ROWS CARDINALITY SELECTIVITY HISTOGRAM NUM_BUCKETS
------------------------------ ---------- ----------- ----------- --------------- -----------
ID 978619 978619 100 NONE 1
MSGID 978619 978619 100 NONE 1
STATUS 978619 1 0 NONE 1
PROCCNT 978619 2 0 NONE 1
ACCEPTTIME 978619 653888 66.82 NONE 1
MSGBODY 978619 0 0 NONE 0
QUEUE 978619 4 0 NONE 1
7 rows selected.
此时又可以走索引:
SQL> explain plan for SELECT messagein0_.proccnt AS col_0_0_ FROM mbfe.messagein messagein0_ where msgID='ID:414d5120514d454d42464520202020204c24b44120263403';
Explained.
SQL> select * from table(dbms_xplan.display());
PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 4204216909
--------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
--------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 55 | 4 (0)| 00:00:01 |
| 1 | TABLE ACCESS BY INDEX ROWID| MESSAGEIN | 1 | 55 | 4 (0)| 00:00:01 |
|* 2 | INDEX RANGE SCAN | INDEXMSGIN_1 | 1 | | 3 (0)| 00:00:01 |
--------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
2 - access("MSGID"='ID:414d5120514d454d42464520202020204c24b44120263403')
14 rows selected.
结论：主键列 唯一列 不需要搜集直方图信息
只有列分布不均匀的时候才需要搜集
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- Oracle直方图导致SQL不走索引.
- 数值转换导致不走索引
- mysql优化，导致查询不走索引的原因总结
- 类型转换导致运行计划不走索引測试案例
- Mysql优化，导致查询不走索引的原因总结
- 解决oracle 11g安装导致数据库无法自动搜集统计信息
- 绑定变量导致执行计划不走索引
- 数据类型不一致导致的SQL不走索引
- 列定义允许null，可能导致Index Hint提升不走索引
- 类型转换导致执行计划不走索引测试案例
- 列定义允许null，可能导致Index Hint提升不走索引
- 数据类型不一致导致的SQL不走索引
- 由于mysql隐式类型转换导致的索引失效的问题
- Oracle执行计划不走索引的原因总结
- oracle 10g 导入到11g中会出现12899错误从而导致数据无法插入，进而数据丢失
- Solr 千万级大数据索引查询导致服务器崩溃的原因和解决
- 中亦科技黄远邦技术人生（15） ——橙色预警：索引空间泄露导致业务中断
- 导致Lucene无法索引查询的NoClassDefFoundError
- 直方图统计导致错误的执行计划
- mysql-不恰当的update语句使用主键和索引导致mysql死锁