Oracle执行计划与统计信息的一些总结

[日期:2011-08-05]
来源:Linux社区  作者:wangshengfeng1986211
[字体: ]

2010-07-01 15:03

1、SET AUTOTRACE ON EXPLAIN
(set autot on exp)
SQLPLUS的命令,在执行SQL语句的同时显示执行计划,设置EXP(LAIN)的目的是只显示执行计划而不显示统计信息.。
2、SQL>explain plan for select ````````;
SQL>select * from table(dbms_xplan.display);

执行了set autotrace on explain语句之后,接下来的查询、插入、更新、删除语句就会显示执行计划,直到执行“set autotrace off;”语句。如果是设置了set autotrace on,除了会显示执行计划之外,还会显示一些有用的统计信息。

执行EXPLAIN PLAN FOR 可以只显示执行计划,然后执行如下查询

SQL> select * from table(dbms_xplan.display);

如:

SQL> explain plan for select * from emp where deptno='20';

Explained.

SQL> select * from table(dbms_xplan.display);

PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------
Plan hash value: 3956160932

--------------------------------------------------------------------------
| Id   | Operation          | Name | Rows   | Bytes | Cost (%CPU)| Time      |
--------------------------------------------------------------------------
|    0 | SELECT STATEMENT   |       |      5 |    150 |      3    (0)| 00:00:01 |
|*   1 |   TABLE ACCESS FULL| EMP   |      5 |    150 |      3    (0)| 00:00:01 |
--------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

PLAN_TABLE_OUTPUT
--------------------------------------------------------------------------------

1 - filter("DEPTNO"=20)

13 rows selected.

3、SQL>exec dbms_stats.delete_table_stats(USER,'表');(删除表的统计信息)

SQL>exec dbms_stats.gather_table_stats(USER,'表',METHOD_OPT=>'FOR ALL COLUMNS SIZE 100')(收集表的统计信息)

4、AUTOTRACE的几个常用选项

set autotrace off ---------------- 不生成autotrace 报告,这是缺省模式
set autotrace on explain ------ autotrace只显示优化器执行路径报告
set autotrace on statistics -- 只显示执行统计信息
set autotrace on ----------------- 包含执行计划和统计信息
set autotrace traceonly ------ 同set autotrace on,但是不显示查询输

(1). set autotrace on explain; --只显示执行计划
SQL> set autotrace on explain;
SQL> 

select count(*) from dba_objects;
COUNT(*)
----------
    31820
Execution Plan
----------------------------------------------------------
  0      SELECT STATEMENT Optimizer=CHOOSE
  1    0   SORT (AGGREGATE)
  2    1     VIEW OF 'DBA_OBJECTS'
  3    2       UNION-ALL
  4    3         FILTER
  5    4           TABLE ACCESS (BY INDEX ROWID) OF 'OBJ$'
  6    5             NESTED LOOPS
  7    6               TABLE ACCESS (FULL) OF 'USER$'
  8    6               INDEX (RANGE SCAN) OF 'I_OBJ2' (UNIQUE)
  9    4           TABLE ACCESS (BY INDEX ROWID) OF 'IND$'
10    9             INDEX (UNIQUE SCAN) OF 'I_IND1' (UNIQUE)
11    3         NESTED LOOPS
12   11           TABLE ACCESS (FULL) OF 'USER$'
13   11           INDEX (RANGE SCAN) OF 'I_LINK1' (NON-UNIQUE)
(2). set autotrace on statistics;--只显示统计信息
SQL> set autotrace on statistics;
SQL> select count(*) from dba_objects;
COUNT(*)
----------
    31820
Statistics
----------------------------------------------------------
         0 recursive calls
         0 db block gets
     25754 consistent gets
         0 physical reads
         0 redo size
       383 bytes sent via SQL*Net to client
       503 bytes received via SQL*Net from client
         2 SQL*Net roundtrips to/from client
         0 sorts (memory)
         0 sorts (disk)
         1 rows processed
(3). set autotrace traceonly;--同set autotrace on 只是不显示查询输出
SQL> set autotrace traceonly;
SQL> select count(*) from dba_objects;
Execution Plan
----------------------------------------------------------
  0      SELECT STATEMENT Optimizer=CHOOSE
  1    0   SORT (AGGREGATE)
  2    1     VIEW OF 'DBA_OBJECTS'
  3    2       UNION-ALL
  4    3         FILTER
  5    4           TABLE ACCESS (BY INDEX ROWID) OF 'OBJ$'
  6    5             NESTED LOOPS
  7    6               TABLE ACCESS (FULL) OF 'USER$'
8    6               INDEX (RANGE SCAN) OF 'I_OBJ2' (UNIQUE)
  9    4           TABLE ACCESS (BY INDEX ROWID) OF 'IND$'
10    9             INDEX (UNIQUE SCAN) OF 'I_IND1' (UNIQUE)
11    3         NESTED LOOPS
12   11           TABLE ACCESS (FULL) OF 'USER$'
13   11           INDEX (RANGE SCAN) OF 'I_LINK1' (NON-UNIQUE)
Statistics
----------------------------------------------------------
         0 recursive calls
         0 db block gets
     25754 consistent gets
         0 physical reads
         0 redo size
       383 bytes sent via SQL*Net to client
       503 bytes received via SQL*Net from client
         2 SQL*Net roundtrips to/from client
         0 sorts (memory)
         0 sorts (disk)
         1 rows processed
(4).set autotrace traceonly explain;--比较实用的选项,只显示执行计划,但是与set autotrace on explain;相比不会执行语句,对于仅仅查看大表的Explain Plan非常管用。
SQL> set autotrace traceonly explain;
SQL> select * from dba_objects;
已用时间: 00: 00: 00.00
Execution Plan
----------------------------------------------------------
  0      SELECT STATEMENT Optimizer=CHOOSE
  1    0   VIEW OF 'DBA_OBJECTS'
  2    1     UNION-ALL
  3    2       FILTER
  4    3         TABLE ACCESS (BY INDEX ROWID) OF 'OBJ$'
  5    4           NESTED LOOPS
  6    5             TABLE ACCESS (FULL) OF 'USER$'
  7    5             INDEX (RANGE SCAN) OF 'I_OBJ2' (UNIQUE)
  8    3         TABLE ACCESS (BY INDEX ROWID) OF 'IND$'
  9    8           INDEX (UNIQUE SCAN) OF 'I_IND1' (UNIQUE)
10    2       TABLE ACCESS (BY INDEX ROWID) OF 'LINK$'
11   10         NESTED LOOPS
12   11           TABLE ACCESS (FULL) OF 'USER$'
13   11           INDEX (RANGE SCAN) OF 'I_LINK1' (NON-UNIQUE)

5、analyze

analyze table hr.employees compute(estimate) statistics;(compute收集每一行数据的统计信息,比较耗时;estimate收集一部分数据行的统计信息)

select t.owner,t.table_name,t.tablespace_name,t.blocks,t.empty_blocks,t.avg_space
from dba_tables t
where t.owner='HR';

上一篇:201521123092《java程序设计》第五周学习总结


下一篇:第一个ExtJS练习(添加用户面板)