index range scan(索引范围扫描):

1.对于unique index来说,如果where 条件后面出现了<,> ,between ...and...的时候,那么就可能执行index range scan,如果where条件后面是=,那么就会执行index unique scan。

2.对于none unique index来说 如果where 条件后面出现了=,>,<,betweed...and...的时候,就有可能执行index range scan。

3.对于组合索引来说,如果where条件后面出现了组合索引的引导列,那么可能执行index range scan。

index fast full scan(索引快速全扫描):

如果select 语句后面中的列都被包含在组合索引中,而且where后面没有出现组合索引的引导列,并且需要检索出大部分数据,那么这个时候可能执行index fast full scan。index fast full scan 发生的条件:

1.必须是组合索引   ?。2.引导列不在where条件中

index skip scan(索引跳跃式扫描)

当查询可以通过组合索引得到结果,而且返回结果很少,并且where条件中没有包含索引引导列的时候,可能执行index skip scan

索引跳跃式扫描发生的条件:

1.必须是组合索引

2.引导列没有出现在where条件中

-eg1

SQL> create table test as select * from dba_objects;

Table created.

SQL> create unique index ind_id on test(object_id);

Index created.

SQL> create index ind_owner on test(owner);

Index created.

SQL> create index ooo on test(owner,object_name,object_type);

Index created.

SQL> exec dbms_stats.gather_table_stats('SCOTT','TEST');

PL/SQL procedure successfully completed.

SQL> set autot trace

SQL> select owner from test where object_id=10;

Execution Plan
----------------------------------------------------------
Plan hash value: 2544773305 --------------------------------------------------------------------------------
------ | Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time
| --------------------------------------------------------------------------------
------ | 0 | SELECT STATEMENT | | 1 | 11 | 2 (0)| 00:0
0:01 | | 1 | TABLE ACCESS BY INDEX ROWID| TEST | 1 | 11 | 2 (0)| 00:0
0:01 | |* 2 | INDEX UNIQUE SCAN | IND_ID | 1 | | 1 (0)| 00:0
0:01 | --------------------------------------------------------------------------------
------ Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"=10) Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
3 consistent gets
0 physical reads
0 redo size
524 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
SQL> select owner from test where object_id<10;

8 rows selected.

Execution Plan
----------------------------------------------------------
Plan hash value: 1361604213 --------------------------------------------------------------------------------
------ | Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time
| --------------------------------------------------------------------------------
------ | 0 | SELECT STATEMENT | | 8 | 88 | 3 (0)| 00:0
0:01 | | 1 | TABLE ACCESS BY INDEX ROWID| TEST | 8 | 88 | 3 (0)| 00:0
0:01 | |* 2 | INDEX RANGE SCAN | IND_ID | 8 | | 2 (0)| 00:0
0:01 | --------------------------------------------------------------------------------
------ Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"<10) Statistics
----------------------------------------------------------
1 recursive calls
0 db block gets
5 consistent gets
0 physical reads
0 redo size
609 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
8 rows processed

对于唯一索引,发生index range scan的时候就是返回多行记录,where后面有<,>,between..and 等返回扫描

SQL> select owner from test where owner='SCOTT';

Execution Plan
----------------------------------------------------------
Plan hash value: 2280863269 ------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 3613 | 21678 | 9 (0)| 00:00:01 |
|* 1 | INDEX RANGE SCAN| IND_OWNER | 3613 | 21678 | 9 (0)| 00:00:01 |
------------------------------------------------------------------------------ Predicate Information (identified by operation id):
--------------------------------------------------- 1 - access("OWNER"='SCOTT') Statistics
----------------------------------------------------------
1 recursive calls
0 db block gets
3 consistent gets
0 physical reads
0 redo size
526 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed

对于非唯一索引,即使where后面是=条件,但也可能返回多行,也是index range scan扫描

SQL> select object_name,object_type from test where owner='SCOTT';

Execution Plan
----------------------------------------------------------
Plan hash value: 2845720098 -------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 3613 | 141K| 28 (0)| 00:00:01 |
|* 1 | INDEX RANGE SCAN| OOO | 3613 | 141K| 28 (0)| 00:00:01 |
------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 1 - access("OWNER"='SCOTT') Statistics
----------------------------------------------------------
1 recursive calls
0 db block gets
4 consistent gets
0 physical reads
0 redo size
610 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed

因为这个索引不是唯一索引,where后面的列用到了索引000,所以进行index range scan

SQL> select owner, object_name,object_type from test where object_name='EMP' ;

no rows selected

Execution Plan
----------------------------------------------------------
Plan hash value: 1799988433 -------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 2 | 80 | 26 (0)| 00:00:01 |
|* 1 | INDEX SKIP SCAN | OOO | 2 | 80 | 26 (0)| 00:00:01 |
------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 1 - access("OBJECT_NAME"='EMP')
filter("OBJECT_NAME"='EMP') Statistics
----------------------------------------------------------
1 recursive calls
0 db block gets
28 consistent gets
0 physical reads
0 redo size
479 bytes sent via SQL*Net to client
513 bytes received via SQL*Net from client
1 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
0 rows processed

上面的查询可以通过索引000来得到,并且where后面没有用到索引列,而且返回的行数很少(。)所以cbo选择index skip scan

select owner, object_name,object_type from test where object_type='INDEX';
Execution Plan
----------------------------------------------------------
Plan hash value: 3464522019 -----------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-----------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1971 | 78840 | 168 (1)| 00:00:03 |
|* 1 | INDEX FAST FULL SCAN| OOO | 1971 | 78840 | 168 (1)| 00:00:03 |
----------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 1 - filter("OBJECT_TYPE"='INDEX') Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
957 consistent gets
0 physical reads
0 redo size
199834 bytes sent via SQL*Net to client
4253 bytes received via SQL*Net from client
341 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
5088 rows processed

同上,但是这里返回行数较多,cbo选择了index fast full scan,避免了全表扫描

Oracle 索引扫描的几种情况的更多相关文章

  1. Oracle 索引扫描的4种类型

    根据索引的类型与where限制条件的不同,有4种类型的Oracle索引扫描: 3,4可归一种 (1) 索引唯一扫描(index uniquescan) (2) 索引范围扫描(index range s ...

  2. Oracle 索引扫描的五种类型

    之前在讨论CBO和RBO的时候提到了索引扫描的几种类型. Oracle Optimizer CBO RBO http://blog.csdn.net/tianlesoftware/archive/20 ...

  3. Oracle索引扫描

    Oracle索引扫描:先通过index查找到索引的值,并根据索引的值对应的rowid值(对于非唯一索引可能返回多个rowid值)直接从表中得到具体的数据.一个rowid唯一的表示一行数据,该行对应的数 ...

  4. MySQL索引失效的几种情况

    1.索引不存储null值 更准确的说,单列索引不存储null值,复合索引不存储全为null的值.索引不能存储Null,所以对这列采用is null条件时,因为索引上根本 没Null值,不能利用到索引, ...

  5. mysql索引总结(4)-MySQL索引失效的几种情况

    mysql索引总结(1)-mysql 索引类型以及创建 mysql索引总结(2)-MySQL聚簇索引和非聚簇索引 mysql索引总结(3)-MySQL聚簇索引和非聚簇索引 mysql索引总结(4)-M ...

  6. oracle数据库中索引失效的几种情况

    原文1:https://blog.csdn.net/u012255097/article/details/102792683 原文2:https://www.cnblogs.com/lanseyita ...

  7. Oracle索引扫描算法

    SQL> create table t as select * from dba_objects; Table created. SQL> create index idx_t on t( ...

  8. oracle 索引扫描类型的分类与构造

    1. INDEX RANGE SCAN--请记住这个INDEX RANGE SCAN扫描方式drop table t purge;create table t as select * from dba ...

  9. mysql 索引失效的几种情况+

随机推荐

  1. uva11827gcd

    gcd裸题,不过输入要注意gets会tle,要用快速读入 #include<map> #include<set> #include<cmath> #include& ...

  2. combobox里面显示checkbox

    看了http://www.cnblogs.com/yubinfeng/p/4463418.html这篇博客,我添加了部分代码,以便在最后获取combobox的value时可以拿到一个数组. HTML代 ...

  3. uva 1511 最小生成树

    https://vjudge.net/problem/UVA-1151 题意,给出N个点以及二维坐标,可以在任意两点间建立通路,代价是两点欧几里得距离的平方,同时有q个套餐,套餐x有qx个点,代价是q ...

  4. IOS-社会化分享

    一.如何实现社交分享 在iOS中,实现“社交分享”的方法 1.自己编写各个平台的分享代码(代码量较多)   2.利用iOS自带的Social.framework   3.利用第三方的分享框架 友盟分享 ...

  5. ARM汇编指令集1

    (汇编)指令是CPU机器指令的助记符,经过编译过会得到一串0011组成的机器码,可以由CPU读取执行. (汇编)伪指令本质不是指令(只是和指令一起写在代码中),它是编译器环境提供的,目的是用来指导编译 ...

  6. Hibernate 悲观锁,乐观锁

    业务逻辑的实现过程中,往往需要保证数据访问的排他性.因此,我们就需要通过一些机制来保证这些数据在某个操作过程中不会被外界修改,这样的机制,在这里,也就是所谓的“锁”,即给我们选定的目标数据上锁,使其无 ...

  7. Intent Flag启动模式P203

    Activity启动模式:点此查看 Intent intent = new Intent(); /** * Intent.FLAG_ACTIVITY_NEW_TASK * 使用一个新的Task来启动一 ...

  8. unity 事件顺序及功能说明

    unity3d中所有控制脚本的基类MonoBehaviour有一些虚函数用于绘制中事件的回调,也可以直接理解为事件函数,例如大家都很清楚的Start,Update等函数,以下做个总结. Awake 当 ...

  9. 【英语】Bingo口语笔记(83) - hell系列

  10. 牛客国庆集训派对Day1:J:Princess Principal(栈模拟求括号匹配)

    题目描述 阿尔比恩王国(the Albion Kingdom)潜伏着一群代号“白鸽队(Team White Pigeon)”的间谍.在没有任务的时候,她们会进行各种各样的训练,比如快速判断一个文档有没 ...