oracle如何降低逻辑读,如何降低该SQL的逻辑读

语句：

SELECT DISTINCT M.PACKAGEID AS PACKAGEID,

M.PROCESSDEFINEID AS PROCESSDEFINEID,

M.WORKFLOWNAME AS WORKFLOWNAME,

B.DOCUMENT_ID AS DOCUMENT_ID,

B.FORM_ID AS FORM_ID,

B.DOCUMENT_TITLE AS DOCUMENT_TITLE,

B.DOCUMENT_PRIORITY AS DOCUMENT_PRIORITY,

TO_CHAR(I1.STARTTIME, 'YYYY-MM-DD HH24:MI:SS') AS STARTTIME,

I1.ACTIVITYNAME AS ACTIVITYNAME,

I1.ACTIVITYID AS ACTIVITYID,

I1.GRANTORID AS GRANTORID,

I1.GRANTORNAME AS GRANTORNAME,

I2.SUBMITUSERNAME AS SUBMITUSERNAME,

BO_INST.BO_INSTANCE_ID AS BO_INSTANCE_ID,

BO_DEF.BO_DEF_NAME AS BO_DEF_NAME,

BO_DEF.BO_DEF_ID AS BO_DEF_ID,

I1.LOGID AS LOGID,

B.DOCUMENT_SECRET AS DOCUMENT_SECRET,

I1.USERNAME AS USERNAME,

ACC.USER_ID AS USERID

FROM BPM70.MOCHA_XJCMCC_ZXCD ZX

INNER JOIN BPM70.MOCHA_DOCUMENT_BODY B ON ZX.DOC_ID = B.BO_INSTANCE_ID

INNER JOIN BPM70.MOCHA_BO_INSTANCE BO_INST

ON BO_INST.BO_INSTANCE_ID = B.BO_INSTANCE_ID

INNER JOIN BPM70.MOCHA_BO_DEFINITION BO_DEF

ON BO_DEF.BO_DEF_ID = BO_INST.BO_DEF_ID

INNER JOIN BPM70.MOCHA_BE_PROCESS_INSTANCE M

ON M.DOCID = BO_INST.BO_INSTANCE_ID

INNER JOIN BPM70.MOCHA_BE_PROCESS_LOG I1

ON I1.PROCESSID = M.PROCESSID

LEFT JOIN BPM70.MOCHA_OSM_ACCOUNTS ACC

ON ACC.ID = I1.USERID

LEFT OUTER JOIN (SELECT LOGID, USERNAME AS SUBMITUSERNAME, PROCESSID

FROM BPM70.MOCHA_BE_PROCESS_LOG) I2

ON I1.FROMLOGID = I2.LOGID

AND I2.PROCESSID = M.PROCESSID

INNER JOIN BPM70.MOCHA_HUMANTASK_TASK LOG_EXT

ON I1.LOGID = LOG_EXT.TASK_ID

WHERE B.DOCUMENT_STATE || '' <> 300

AND I1.WORK_STATUS_ID = 2

AND BO_DEF.STATE || '' IN (1, 3)

AND BO_INST.SUBMIT_FLAG = 1

AND BO_INST.SAVE_FLAG = 1

AND GENTIME > TO_DATE(:1, 'YYYY-MM-DD HH24:MI:SS')

执行计划：

PLAN_TABLE_OUTPUT

------------------------------------------------------------------------------------------------------------------------------------------------------

---------------------------------------------------------------------------------------------------

---------------------------------------------------------------------------------------------------

| 0 | SELECT STATEMENT | | 21 | 10836 | 3841 |

| 1 | SORT UNIQUE | | 21 | 10836 | 3841 |

| 2 | NESTED LOOPS OUTER | | 21 | 10836 | 3824 |

| 3 | NESTED LOOPS | | 21 | 10353 | 3782 |

| 4 | VIEW | | 21 | 10227 | 3761 |

| 5 | NESTED LOOPS OUTER | | 21 | 15036 | 3761 |

| 6 | NESTED LOOPS | | 21 | 13965 | 3740 |

| 7 | NESTED LOOPS | | 18 | 9882 | 3668 |

| 8 | HASH JOIN | | 24 | 8880 | 3596 |

| 9 | TABLE ACCESS FULL | MOCHA_XJCMCC_ZXCD | 884 | 42432 | 19 |

| 10 | HASH JOIN | | 429 | 134K| 3576 |

| 11 | TABLE ACCESS BY INDEX ROWID| MOCHA_BO_INSTANCE | 1299 | 109K| 566 |

| 12 | NESTED LOOPS | | 4110 | 606K| 1700 |

| 13 | TABLE ACCESS FULL | MOCHA_BO_DEFINITION | 3 | 195 | 2 |

| 14 | INDEX RANGE SCAN | IDX_BO_INST_DEF_ID | 2732 | | 12 |

| 15 | TABLE ACCESS FULL | MOCHA_DOCUMENT_BODY | 15872 | 2650K| 1856 |

| 16 | TABLE ACCESS BY INDEX ROWID | MOCHA_BE_PROCESS_INSTANCE | 1 | 179 | 3 |

| 17 | INDEX RANGE SCAN | IDX_PROI_DOCID | 1 | | 2 |

| 18 | TABLE ACCESS BY INDEX ROWID | MOCHA_BE_PROCESS_LOG | 1 | 116 | 4 |

| 19 | INDEX RANGE SCAN | IDX_PROLOG_PID | 4 | | 2 |

| 20 | TABLE ACCESS BY INDEX ROWID | MOCHA_OSM_ACCOUNTS | 1 | 51 | 1 |

| 21 | INDEX UNIQUE SCAN | PK_MOCHA_OSM_ACCOUNTS | 1 | | |

| 22 | INDEX UNIQUE SCAN | PK_HUMANTASK_TASK | 1 | 6 | 1 |

| 23 | TABLE ACCESS BY INDEX ROWID | MOCHA_BE_PROCESS_LOG | 1 | 23 | 2 |

| 24 | INDEX UNIQUE SCAN | PK_BE_PROCESS_LOG | 1 | | 1 |

---------------------------------------------------------------------------------------------------

SQL> select owner,partitioned,last_analyzed,num_rows from dba_tables where table_name='MOCHA_XJCMCC_ZXCD';

OWNER PARTIT LAST_ANAL NUM_ROWS

------------------------------------------------------------ ------ --------- ----------

BPM70 NO 16-APR-11 17680

SQL> select owner,partitioned,last_analyzed,num_rows from dba_tables where table_name='MOCHA_DOCUMENT_BODY';

OWNER PARTIT LAST_ANAL NUM_ROWS

------------------------------------------------------------ ------ --------- ----------

BPM70 NO 16-APR-11 31,7443

表 MOCHA_DOCUMENT_BODY各个列数据的分布程度：

COLUMN_NAME NUM_DISTINCT

-------------------- ------------

ORG_ALL_PATH 1

START_TIME 317438

DOCUMENT_SECRET 1

SPREAD1 0

SPREAD2 0

SPREAD3 0

SPREAD4 0

SPREAD5 46668

SPREAD6 0

SPREAD7 310

SPREAD8 116

SPREAD9 0

SPREAD10 0

SPREAD11 0

SPREAD12 0

SPREAD13 0

SPREAD14 0

SPREAD15 0

SPREAD16 0

SPREAD17 0

SPREAD18 0

SPREAD19 0

SPREAD20 0

DOCUMENT_ID 317443

BO_INSTANCE_ID 317443

FORM_ID 150

FORM_VERSION 36

DOCUMENT_STATE 1

DOCUMENT_PRIORITY 4

DOCUMENT_TITLE 28220

ORG_ID 325

DOC_HTML_HEIGHT 1

MODIFYTIME 317443

ORG_PATH 1

SQL> select index_name,column_name,column_position from dba_ind_columns where table_name='MOCHA_DOCUMENT_BODY';

INDEX_NAME COLUMN_NAME COLUMN_POSITION

------------------------------ ------------------------- ---------------

PK_DOCUMENT DOCUMENT_ID 1

IDX_BODY_INST_ID BO_INSTANCE_ID 1

IDX_DOC_BODY_DOC_TITLE DOCUMENT_TITLE 1

Statistics

----------------------------------------------------------

0 recursive calls

0 db block gets

42,5057 consistent gets

16 physical reads

0 redo size

19337 bytes sent via SQL*Net to client

914 bytes received via SQL*Net from client

10 SQL*Net roundtrips to/from client

1 sorts (memory)

0 sorts (disk)

132 rows processed

请教如何降低该逻辑读？？？？？？

oracle如何降低逻辑读,如何降低该SQL的逻辑读相关推荐

Interview：算法岗位面试—10.24下午—上海某软件公司(机器学习，上市)电话面试—考察SVM、逻辑回归、降低过拟合、卷积网络基础等
Interview:算法岗位面试-10.24下午-上海某软件公司(机器学习,上市)电话面试-考察SVM.逻辑回归.降低过拟合.卷积网络基础等导读:当时电话来的非常快,我刚做完一家公司的笔试,接着来了 ...
oracle怎么判断逻辑读,如何查看Oracle数据库物理读、逻辑读前10的sql？
概述今天主要介绍几个sql,可能很多人就收藏一下但很少会去用,所以结合案例来做一下分析,这样就知道在什么场景用了. 下面介绍一下案例. 1.监控分析从监控发现物理IO读处于比较高的水平,截图如下: ...
初谈SQL Server逻辑读、物理读、预读
前言本文涉及的内容均不是原创,是记录自己在学习IO.执行计划的过程中学习其他大牛的博客和心得并记录下来,之所以想写下来是为了记录自己在追溯的过程遇到的几个问题,并把这些问题弄清楚. 本章最后已贴出原 ...
MySQL预读失效_华为云MySQL新增“逻辑预读”特性，轻松解决线性预读失效问题...
随着用户对数据访问速度的日益重视,MySQL数据库在最初的设计中,采用了线性预读的方式,提前将即将使用的数据预读到Buffer pool中,来提升数据的访问速度,但在实际使用过程中,线性预读失效的问题 ...
浅谈SQL Server逻辑读、物理读、预读
目录 (一)前言 (二)正文 1. SQL Server的数据存储方式 2. 剖析IO统计信息 (1)初识三读 (2)逻辑读.物理读.预读的关系 (3)看图理解[逻辑读].[物理读].[预读] (三) ...
oracle dg3种模式,DG有下面三种模式逻辑standby和物理standby
DG有下面三种模式 – Maximum protection – Maximum availability – Maximum performance 在Maximum protection下, 可以 ...
逻辑回归三部曲——逻辑回归(logistics regression)原理-让你彻底读懂逻辑回归
逻辑回归已经在各大银行和公司都实际运用于业务,已经有很多前辈写过逻辑回归.本文将从我实际应用的角度阐述逻辑回归原理,致力于让逻辑回归变得清晰.易懂.逻辑回归又叫对数几率回归,是一种广义的线性 ...
状态机复杂逻辑问题_状态管理中的逻辑
状态机复杂逻辑问题 by Oguz Gelal 由Oguz Gelal 状态管理中的逻辑 (The Logic in State Management) The standardization of ...
商业逻辑12讲之技术创新的逻辑
目录 1. 概述 2. 创新的重要性 3. 创新的基本概念 4. 创新的管理模块 5. 技术创新的战略管理 6. 技术创新的组织设计 7. 培养创新人才 8. 技术创新的文化建设 9. 小结 1. 概 ...
数据库设计读多写少、写多读少、写多读多各场景数据库建设方案
数据库建设方案读多写少写多读少写多读多数据库集群方案优缺点读多写少解决方案:采用传统关系型数据库足以应对,若并发量很大,采用mysql集群即可应对! 写多读少 1.业务场景:滴滴.饭堂刷卡 ...

oracle如何降低逻辑读,如何降低该SQL的逻辑读

oracle如何降低逻辑读,如何降低该SQL的逻辑读相关推荐

最新文章

热门文章