Oracle 自动生成的视图VM_NSO_1
有时候优化sql的时候，在执行计划中看到有VM_NSO_X的视图，在Oracle定义中，可以吧NSO理解为nested subquery optimizing，功能就是把in转换为join，把not in转换为anti join等，当然转换的时候有一定的限制。下面我们来简单看下会生成VM_NSO_1视图的几个例子
1. 创建2个测试表
SQL> create table test_jerry as select * from dba_objects; Table created. SQL> create table test_jerry2 as select * from dba_objects; Table created. SQL> select count(*) from test_jerry where object_id not in (select max(object_id) from test_jerry2 group by owner); Execution Plan
----------------------------------------------------------
Plan hash value: 3525080607 ------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time |
------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 26 | | 908 (1)| 00:00:11 |
| 1 | SORT AGGREGATE | | 1 | 26 | | | |
|* 2 | HASH JOIN RIGHT ANTI NA| | 89846 | 2281K| 2144K| 908 (1)| 00:00:11 |
| 3 | VIEW | VW_NSO_1 | 87509 | 1110K| | 349 (1)| 00:00:05 |
| 4 | HASH GROUP BY | | 87509 | 2563K| | 349 (1)| 00:00:05 |
| 5 | TABLE ACCESS FULL | TEST_JERRY2 | 87509 | 2563K| | 347 (1)| 00:00:05 |
| 6 | TABLE ACCESS FULL | TEST_JERRY | 89847 | 1140K| | 347 (1)| 00:00:05 |
------------------------------------------------------------------------------------------------ Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"="MAX(OBJECT_ID)") Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
2490 consistent gets
2484 physical reads
0 redo size
528 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 count(*) from test_jerry where object_id not in (select max(object_id) from test_jerry2); Execution Plan
----------------------------------------------------------
Plan hash value: 3071647562 ------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 13 | 694 (1)| 00:00:09 |
| 1 | SORT AGGREGATE | | 1 | 13 | | |
|* 2 | TABLE ACCESS FULL | TEST_JERRY | 4492 | 58396 | 347 (1)| 00:00:05 |
| 3 | SORT AGGREGATE | | 1 | 13 | | |
| 4 | TABLE ACCESS FULL| TEST_JERRY2 | 87509 | 1110K| 347 (1)| 00:00:05 |
------------------------------------------------------------------------------------ Predicate Information (identified by operation id):
--------------------------------------------------- 2 - filter("OBJECT_ID"<> (SELECT MAX("OBJECT_ID") FROM "TEST_JERRY2"
"TEST_JERRY2")) Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
7 recursive calls
0 db block gets
2629 consistent gets
2484 physical reads
0 redo size
528 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 count(*) from test_jerry where object_id in (select object_id from test_jerry2 where owner='SYS' minus select object_id from test_jerry where owner='SCOTT'); Execution Plan
----------------------------------------------------------
Plan hash value: 773093838 ----------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time |
----------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 26 | | 1307 (1)| 00:00:16 |
| 1 | SORT AGGREGATE | | 1 | 26 | | | |
|* 2 | HASH JOIN | | 32153 | 816K| | 1307 (1)| 00:00:16 |
| 3 | VIEW | VW_NSO_1 | 32153 | 408K| | 960 (1)| 00:00:12 |
| 4 | MINUS | | | | | | |
| 5 | SORT UNIQUE | | 32153 | 941K| 1272K| | |
|* 6 | TABLE ACCESS FULL| TEST_JERRY2 | 32153 | 941K| | 347 (1)| 00:00:05 |
| 7 | SORT UNIQUE | | 14 | 420 | | | |
|* 8 | TABLE ACCESS FULL| TEST_JERRY | 14 | 420 | | 347 (1)| 00:00:05 |
| 9 | TABLE ACCESS FULL | TEST_JERRY | 89847 | 1140K| | 347 (1)| 00:00:05 |
---------------------------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"="OBJECT_ID")
6 - filter("OWNER"='SYS')
8 - filter("OWNER"='SCOTT') Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
93 recursive calls
0 db block gets
4691 consistent gets
3726 physical reads
0 redo size
528 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
12 sorts (memory)
0 sorts (disk)
1 rows processed SQL> select count(*) from test_jerry where object_id in (select object_id from test_jerry2 where owner='SYS' union all select object_id from test_jerry where owner='SCOTT'); Execution Plan
----------------------------------------------------------
Plan hash value: 1173723582 --------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
--------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 26 | 1041 (1)| 00:00:13 |
| 1 | SORT AGGREGATE | | 1 | 26 | | |
|* 2 | HASH JOIN | | 32167 | 816K| 1041 (1)| 00:00:13 |
| 3 | VIEW | VW_NSO_1 | 32167 | 408K| 694 (1)| 00:00:09 |
| 4 | HASH UNIQUE | | 32167 | 942K| 694 (1)| 00:00:09 |
| 5 | UNION-ALL | | | | | |
|* 6 | TABLE ACCESS FULL| TEST_JERRY2 | 32153 | 941K| 347 (1)| 00:00:05 |
|* 7 | TABLE ACCESS FULL| TEST_JERRY | 14 | 420 | 347 (1)| 00:00:05 |
| 8 | TABLE ACCESS FULL | TEST_JERRY | 89847 | 1140K| 347 (1)| 00:00:05 |
-------------------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"="OBJECT_ID")
6 - filter("OWNER"='SYS')
7 - filter("OWNER"='SCOTT') Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
82 recursive calls
0 db block gets
4669 consistent gets
3726 physical reads
0 redo size
527 bytes sent via SQL*Net to client
524 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
8 sorts (memory)
0 sorts (disk)
1 rows processed
SQL> select count(*) from test_jerry where object_id in (SELECT LEVEL FROM DUAL CONNECT BY LEVEL<100); Execution Plan
----------------------------------------------------------
Plan hash value: 3708743834 --------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
--------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 26 | 350 (1)| 00:00:05 |
| 1 | SORT AGGREGATE | | 1 | 26 | | |
|* 2 | HASH JOIN | | 1 | 26 | 350 (1)| 00:00:05 |
| 3 | VIEW | VW_NSO_1 | 1 | 13 | 3 (34)| 00:00:01 |
| 4 | HASH UNIQUE | | 1 | | 3 (34)| 00:00:01 |
|* 5 | CONNECT BY WITHOUT FILTERING (UNIQUE)| | | | | |
| 6 | FAST DUAL | | 1 | | 2 (0)| 00:00:01 |
| 7 | TABLE ACCESS FULL | TEST_JERRY | 89847 | 1140K| 347 (1)| 00:00:05 |
-------------------------------------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 2 - access("OBJECT_ID"="LEVEL")
5 - filter(LEVEL<100) Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
4 recursive calls
0 db block gets
1315 consistent gets
1242 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
1 sorts (memory)
0 sorts (disk)
1 rows processed
可以从上面的sql得出一个简单的结论，当子查询中出现max，rownum，group by，union all，minus，intersect等聚合函数的时候，Oracle就会自动把子查询转换成视图VM_NSO_X，其实在Oracle的子查询中如果出现上面的几种情况，也就限制了view merge，就无法对视图进行merge。如果需要去掉CBO生成VM_NSO_1，只需要在子查询中加一个no_unnest限制，这样CBO就会走filter了。下面看一个加了no_unnest的执行计划
SQL> select count(*) from test_jerry where object_id in (SELECT /*+ no_unnest() */ LEVEL FROM DUAL t CONNECT BY LEVEL<100); Execution Plan
----------------------------------------------------------
Plan hash value: 2000702637 -------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 13 | 5963 (1)| 00:01:12 |
| 1 | SORT AGGREGATE | | 1 | 13 | | |
|* 2 | FILTER | | | | | |
| 3 | TABLE ACCESS FULL | TEST_JERRY | 89847 | 1140K| 347 (1)| 00:00:05 |
|* 4 | FILTER | | | | | |
|* 5 | CONNECT BY WITHOUT FILTERING (UNIQUE)| | | | | |
| 6 | FAST DUAL | | 1 | | 2 (0)| 00:00:01 |
------------------------------------------------------------------------------------------------------- Predicate Information (identified by operation id):
--------------------------------------------------- 2 - filter( EXISTS (SELECT /*+ NO_UNNEST */ 0 FROM "SYS"."DUAL" "T" WHERE LEVEL=:B1 CONNECT
BY LEVEL<100))
4 - filter(LEVEL=:B1)
5 - filter(LEVEL<100) Note
-----
- dynamic sampling used for this statement (level=2) Statistics
----------------------------------------------------------
7 recursive calls
0 db block gets
1385 consistent gets
1242 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
86968 sorts (memory)
0 sorts (disk)
1 rows processed
- 本文固定链接: http://www.savedba.com/?p=824
- 转载请注明: 版权所有，文章允许转载，但必须以链接方式注明源地址，否则追究法律责任!
Oracle 自动生成的视图VM_NSO_1的更多相关文章
- Oracle 自动生成hive建表语句
从 oracle 数据库导数到到 hive 大数据平台,需要按照大数据平台的数据规范,重新生成建表的 SQL 语句,方便其间,写了一个自动生成SQL的存储过程. ① 创建一张表,用来存储源表的结构,以 ...
- oracle自动编号
oracle自动编号 在access中有自动编号的数据类型,MSSQL和MYSQL也都有自动增长的数据类型,插入记录时不用操作此字段,会自动获得数据值,而oracle没有自动增长的数据类型,我们需要建 ...
- ORACLE 创建与使用视图
一.what(什么是视图?) 1.视图是一种数据库对象,是从一个或者多个数据表或视图中导出的虚表,视图所对应的数据并不真正地存储在视图中,而是存储在所引用的数据表中,视图的结构和数据是对数据表进行查询 ...
- Oracle 创建和使用视图
一.what(什么是视图?) 1.视图是一种数据库对象,是从一个或者多个数据表或视图中导出的虚表,视图所对应的数据并不真正地存储在视图中,而是存储在所引用的数据表中,视图的结构和数据是对数据表进行查询 ...
- 利用 Oracle EM 企业管理器 进行oracle SQL的优化（自动生成索引）
利用 Oracle EM 企业管理器 进行oracle SQL的优化(自动生成索引) ##应用情景 项目中有大量的SQL,尤其是涉及到统计报表时,表关联比较多,当初开发建表时也没搞好索引关联的,上线后 ...
- 基于OCILIB的oracle数据库操作总结及自动生成Model和Dao的工具
基于OCILIB的oracle数据库操作总结 1. 类图 2. 基类BaseOCIDao的设计与实现 BaseOCIDao.h头文件 #pragma once /* ----- ...
- 简单两步快速学会使用Mybatis-Generator自动生成entity实体、dao接口和简单mapper映射（用mysql和oracle举例）
前言: mybatis-generator是根据配置文件中我们配置的数据库连接参数自动连接到数据库并根据对应的数据库表自动的生成与之对应mapper映射(比如增删改查,选择性增删改查等等简单语句)文件 ...
- Oracle 每天自动生成AWR报告
经验丰富的老员工希望能够每天为数据库生成1个AWR报告,以便于后期分析数据库的性能变化,手动生成太麻烦,查了一下资料,发现可以自动生成,过程如下. 数据库环境:11gR2 RAC(双节点) AWR报告 ...
- Oracle 函数 “自动生成订单号”
create or replace function get_request_code return varchar2 AS --函数的作用:自动生成订单号 v_mca_no mcode_apply_ ...
随机推荐
- nginx配置白名单
配置如下: http模块: http { include mime.types; default_type application/octet-stream; #log_format main '$r ...
- C# 3.0 扩展方法&接口
namespace ExtensionInterfaceMethod { class Program { static void Main(string[] args) { //使用接口变量来调用扩展 ...
- contextMenuEvent
#include "mainwindow.h" #include "ui_mainwindow.h" #include <QDesktopWidget&g ...
- 【一起来烧脑】一步学会JavaScript体系
[外链图片转存失败(img-b0GOhxRY-1563571645197)(https://upload-images.jianshu.io/upload_images/11158618-ba249b ...
- UOJ310. 【UNR #2】黎明前的巧克力 [FWT]
UOJ 思路 显然可以转化一下,变成统计异或起来等于0的集合个数,这样一个集合的贡献是\(2^{|S|}\). 考虑朴素的\(dp_{i,j}\)表示前\(i\)个数凑出了\(j\)的方案数,发现这其 ...
- Django基础（1）-虚拟环境的安装及配置
virtualenv介绍 (1)做什么的?virtualenv是用于创建独立的python环境,使得多个python应用彼此独立: (2)优点: a)使不同应用开发环境独立 b)环境升级不影响其他应用 ...
- mac eclipse 创建Java 工程
首先创建Java工程testjavapro,创建包testjavapro,接着创建类testjava 参考: https://www.jianshu.com/p/20280b850c95
- python定制后处理云图
用后处理软件处理的云图会出现这样或那样的不满意,其实我们可以将求解数据导出以后,借助python定制云图. 我们以fluent为例 求解完成之后,我们将我们需要做云图的物理量以ASCII导出 如下的p ...
- JavaWeb之问题集(1) —— Tomcat启动闪退
1. 问题描述 环境: System:Windows 10 Professional JDK:1.8.0.212 Tomcat:9.0.22 原本可以正常的启动和关闭,后来突然的就无法使用了,无法启动 ...
- Perl看完这个,再不敢说自己会玩贪吃蛇
某天闲逛时看见一副动图: 真的是非常贪吃,各种拐弯各种吃,感觉十分有趣. 用Perl来实现自动吃满,蓄谋已久,之前的字符贪吃蛇.深度优先算法.A*算法,都是为此篇做铺垫. 那么,怎样让蛇不吃到自己呢? ...