关于软解析与硬解析,以及session cached cursors (翻译,ref asktom) (摘自CNOUG)

翻译 2004年10月11日 09:27:00
quote from difference between soft parse and hard parse

-- Thanks for the question regarding "difference between soft parse and hard parse", version
originally submitted on 26-Jan-2002 19:17 Eastern US time, last updated 5-Aug-2004 13:07

Hi Tom

Can you explain briefly the difference between soft and hard parse?


and we said...

Here is a long winded answer, it is extracted in part from a new book oming out soon "beginning Oracle programming" that I collaborated on:


This is the first step in the processing of any statement in Oracle.  Parsing is
the act of breaking the submitted statement down into its component parts ? determining what type of statement it is (query, DML, DDL) and performing various checks on it.  

The parsing process performs two main functions:

Syntax Check: is the statement a valid one?  Does it make sense given the SQL grammar documented in the SQL Reference Manual.  Does it follow all of the rules for SQL.
1. 语法检查,这条语句是否合理? 它是否符合SQL语法文档<>的相关语法,是不是符合SQL语句的所有规则.

o Semantic Analysis:  Going beyond the syntax ? is the statement valid in light of the objects in the database (do the tables and columns referenced exist).  Do you have access to the objects ? are the proper privileges in place?  Are there ambiguities in the statement ? for example if there are two tables T1 and T2 and both have a column X, the query ?select X from T1, T2 where ?? is ambiguous, we don?t know which table to get X from.  And so on.
2.语义分析: 已经处理完语法检查了吗? 这条语句是否正确的引用了数据库中对象
(是否语句中所有的表和字段都存在). 是否有访问这些对象的权限? 是不是对应的权限都有效(主要指role对应的权限吧,biggrin.gif)? 比如是否有如下的两个表T1,T2,而这两个表有相同的字段名column X,而查询语句 ? select X from T1,T2 where ??,(没有明确指定列名),我们无法知道从哪个表去取出字段X的值,以及类似的一系列问题.

So, you can think of parsing as basically a two step process, that of a syntax
check to check the validity of the statement and that of a semantic check ? to ensure the statement can execute properly.  The difference between the two types of checks are hard for you to see ? Oracle does not come back and say ?it failed the syntax check?, rather it returns the statement with a error code and message.  So for example, this statement fails with a syntax error:
因此,你可以认为解析基本上是由这两个步骤组成的: 检查语句有效性的语法检查和确保语句能够正常运行的语义检查.这两种检查的区别就是你所要说的硬解析了.Oracle不会特别的指出这条语句没有通过语法检查,它给你返回这条语句对应的错误代码和错误信息. 比如: 下面这条语句没有通过语法检查.

  Code: [Copy to clipboard]  
SQL> select from where 2;
select from where 2
ERROR at line 1:
ORA-00936: missing expression

While this statement failed with a semantic error ? if the table NOT_A_TABLE existed and we had permission to access it, this statement would have succeeded:

而这条语句没有通过语义检查? 如果表NOT_A_TABLE存在,并且我们具有访问这个表的权限,这条语句就会正确执行了.

  Code: [Copy to clipboard]  
SQL> select * from not_a_table;
select * from not_a_table
ERROR at line 1:
ORA-00942: table or view does not exist

That is the only way to really tell the difference between a semantic and
syntactic error ? if the statement COULD have executed given the proper objects and privileges, you had a semantic error, otherwise if the statement could not execute under any circumstances, you have a syntax error.  Regardless ? Oracle will not execute the statement for you!


The next step in the parse operation is to see if the statement we are currently parsing has already in fact been processed by some other session.  If it has ? we may be in luck here, we can skip the next two steps in the process, that of optimization and row source generation.  If we can skip these next two steps in the process, we have done what is known as a Soft Parse ? a shorter process to getting our query going.  If we cannot, if we must do all of the steps, we are performing what is known as a Hard Parse ? we must parse, optimize, generate the plan for the query.  This distinction is very important.  When developing our applications we want a very high percentage of our queries to be Soft Parsed ? to be able to skip the optimize/generate phases ? as they are very CPU intensive as well as a point of contention (serialization).  If we have to Hard Parse a large percentage of our queries, our system will function slowly and in some cases ? not at all.

解析中的下一个步骤就是检查我们现在正在解析的语句是否已经被其他的session执行过,在这一步我们可能很幸运,我们可以跳过下面的两个步骤: 语句的优化(生成执行计划) 和 生成执行编码. 如果我们能够跳过这两个步骤,我们就实现了一般意义上的软解析, 我们必须解析,优化,并为这条语句生成执行计划.这个特性(distinction)非常重要,当我们开发应用程序的时候,我们会希望大部分的语句都是软解析的,以跳过优化和生成编码的步骤,因为他们和争用(串行化,)(contention)一样,都是非常消耗CPU的操作.如果我们必须硬解析我们语句的大部分的话,在某些情况下我们的系统就会运行的很慢,并不是任何时候(在olap,dss系统应该是个例外吧smile.gif)

The way this sharing of SQL in Oracle is accomplished is via the shared pool, a piece of memory in the SGA maintained by Oracle.  We covered this topic in chapter 5 but will revisit it again in the context of processing a query.  After Oracle parses the query and it passes the syntax and semantic checks ? it will look in the shared pool component of the SGA to see if that same exact query has already been processed by another session.  Since it has performed the semantic check it has already figured out:

在Oracle中共享SQL的方式是通过共享池--SGA中的一块内存(由Oracle系统来维护)--实现的.我们在第五章中讨论了这个主题,并且会在查询处理的章节再次讨论这个主题.当Oracle已经解析了这条语句,并且已经通过语法和语义检查的时候,Oracle就会在SGA的shared pool(共享池)组件中查询,看看,是否已经有一个完全一样的语句已经被另外一个session执行过.因为语句已经通过了oracle的语义检查,并且oracle已经计算出:

o Exactly what tables are involved

o That we have access to the tables (the proper privileges are there)

And so on.  Now, it can look at all of the queries in the shared pool that have already been parsed/optimized and generated to see if the work has already been done.  

Soft Parse and session_cashed_cursor parameter  January 28, 2002
Reviewer:  A reader


can you explain what means the default value (0) of session_ casched_cursor parameter? Oracle always has a cache of sqls in sga and size of this cache is determinated through init.ora parameter shared_pool_size.

tom : 你能不能给我解释一下,session_cashed_cursor(默认值为0)参数的含义,Oracle在SGA中始终保持一个sql语句的缓存.这个缓存的大小是由初始化参数shared_pool_size的大小决定的.

The default value of ZERO means that by default, ZERO cursors will be cached for your session.


They will be cached in the shared pool -- but your session will have to find
them there.  Session Cached Cursors can remove the need to have to "find" them.  Consider this example that shows the difference spent when soft parsing without session cached cursors vs soft parsing WITH session cached cursors:
他们会被缓存在共享池(Shared Pool)中,但是你的session必须在共享池中去查找他们,session cached cursors可以省略掉去共享池中查找这一步.下面这个例子

  Code: [Copy to clipboard]  
ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> create table emp as select * from scott.emp;
Table created.

ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> create table run_stats ( runid varchar2(15),
name varchar2(80), value int );
Table created.

ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> create or replace view stats
  2  as select 'STAT...' || a.name name, b.value
  3        from v$statname a, v$mystat b
  4       where a.statistic# = b.statistic#
  5      union all
  6      select 'LATCH.' || name,  gets
  7        from v$latch;

View created.

ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> column name format a40
ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> declare
  2      l_start number;
  3          l_cnt   number;
  4  begin
  5      execute immediate 'alter session set session_cached_cursors=0';
  6      insert into run_stats select 'before', stats.* from stats;
  8      l_start := dbms_utility.get_time;
  9      for i in 1 .. 1000
10      loop
11            execute immediate 'select count(*) from emp' into l_cnt;
12      end loop;
13      dbms_output.put_line( (dbms_utility.get_time-l_start) || ' hsecs' );
15      execute immediate 'alter session set session_cached_cursors=100';
16      insert into run_stats select 'after 1', stats.* from stats;
18      l_start := dbms_utility.get_time;
19      for i in 1 .. 1000
20      loop
21            execute immediate 'select count(*) from emp' into l_cnt;
22      end loop;
23      dbms_output.put_line( (dbms_utility.get_time-l_start) || ' hsecs' );
25      insert into run_stats select 'after 2', stats.* from stats;
26  end;
27  /
45 hsecs
35 hsecs

PL/SQL procedure successfully completed.

so, session cached cursors RAN faster (i ran this a couple of times, there
were no hard parses going on.  But the real good news is:


  Code: [Copy to clipboard]  
ops$[email]tkyte@ORA817DEV.US.ORACLE.COM[/email]> select a.name, b.value-a.value run1,
c.value-b.value run2,
  2         ( (c.value-b.value)-(b.value-a.value)) diff
  3    from run_stats a, run_stats b, run_stats c
  4   where a.name = b.name
  5     and b.name = c.name
  6     and a.runid = 'before'
  7     and b.runid = 'after 1'
  8     and c.runid = 'after 2'
  9     and (c.value-a.value) > 0
10     and (c.value-b.value) <> (b.value-a.value)
11   order by abs( (c.value-b.value)-(b.value-a.value))
12  /

NAME                                           RUN1       RUN2       DIFF
---------------------------------------- ---------- ---------- ----------
LATCH.checkpoint queue latch                      3          4          1
LATCH.redo allocation                            30         31          1
STAT...consistent gets                         5088       5089          1
STAT...deferred (CURRENT) block cleanout          2          3          1

STAT...calls to get snapshot scn: kcmgss       5019       5018         -1
STAT...enqueue releases                          10          9         -1
STAT...execute count                           1015       1014         -1
STAT...opened cursors cumulative               1015       1014         -1
STAT...parse count (total)                     1015       1014         -1
STAT...session cursor cache count                 0          1          1
STAT...redo entries                              28         27         -1
STAT...recursive calls                         1180       1179         -1
STAT...physical reads                             1          0         -1
LATCH.direct msg latch                            2          0         -2
LATCH.session queue latch                         2          0         -2
LATCH.done queue latch                            2          0         -2
STAT...free buffer requested                      8          6         -2
STAT...enqueue requests                          11          9         -2
LATCH.messages                                    3          0         -3
STAT...db block changes                          47         44         -3
LATCH.redo writing                                3          0         -3
LATCH.ksfv messages                               4          0         -4
STAT...session logical reads                  17128      17123         -5
LATCH.row cache objects                         184        178         -6
STAT...db block gets                          12040      12034         -6
STAT...parse time elapsed                         9          3         -6
STAT...parse time cpu                            13          4         -9
STAT...recursive cpu usage                       51         38        -13
LATCH.cache buffers chains                    34315      34335         20
STAT...redo size                              23900      24000        100
STAT...session cursor cache hits                  3       1002        999
LATCH.shared pool                              2142       1097      -1045
LATCH.library cache                           17361       2388     -14973

34 rows selected.


see the significantly REDUCED number of LATCH counts on the library and shared pool.  Since a latch is a lock, a lock is a serialization device, serialization
implies WAITS -- using the session cached cursors will increase scalability and performance as you add more and more users.  Its not only FASTER, but more scalable as well...




转载于http://www.cnblogs.com/eastsea/p/3904874.html 2014-08-11 Created By BaoXinjian 一、摘要 Orac...
  • qq_24369113
  • qq_24369113
  • 2016年08月05日 16:33
  • 765


Oracle数据库中的CURSOR分为两种类型:Shared Cursor 和 Session Cursor 1,Shared Cursor Oracle里的第一种类型的Cursor就是Shared ...
  • waterxcfg304
  • waterxcfg304
  • 2016年06月25日 14:03
  • 1572


硬解析和软解析有相同的一步,而软软解析与硬解析、软解析完全不一样。先来说下理论上的东西,然后来做个实验。 硬解析过程:     1.语法、语义及权限检查;     2.查询转换(通过应用各种不同的转换...
  • guogang83
  • guogang83
  • 2013年12月11日 17:47
  • 5537

oracle学习笔记 SQL语句解析类型_硬解析_软解析

oracle学习笔记 SQL语句解析类型_硬解析_软解析 shared pool在讲oracle体系结构概述时讲了大体是干什么的, 这节把一些细节讲一下。 一)软件出问题 在...
  • xinzhan0
  • xinzhan0
  • 2016年07月28日 17:29
  • 778

Oracle 硬解析和软解析

以下内容转自网络: ------------------------------------- 说到软解析(soft prase)和硬解析(hard prase),就不能不说一下Oracl...
  • IndexMan
  • IndexMan
  • 2013年04月12日 11:47
  • 1431

Oracle 硬解析与软解析

--======================= -- Oracle 硬解析与软解析 --======================= Oracle 硬解析与软解析是我们经常遇到的问题,什么情况会...
  • fuzhangpeng
  • fuzhangpeng
  • 2013年09月08日 17:40
  • 434


Oracle的硬解析和软解析 提到软解析(soft prase)和硬解析(hard prase),就不能不说一下Oracle对sql的处理过程。当你发出一条sql语句交付Oracle,在执...
  • holandstone
  • holandstone
  • 2015年11月30日 10:04
  • 3544


本节讲解shared pool的内存结构   SQL语句的执行分为三个阶段: 1.      SQL的解析(parse); 2.      SQL的执行(exec); 3.      SQL...
  • chengonghao
  • chengonghao
  • 2016年03月21日 19:12
  • 481

共享池之八:软解析、硬解析、软软解析 详解一条SQL在library cache中解析涉及的锁

先来张大图: 结合上图来说明一下解析的各个步骤涉及的锁。 软解析、硬解析、软软解析区别的简单说明: 为了将用户写的sql文本转化为oracle认识的且可执行的语句,这个过程就叫做解析...
  • q947817003
  • q947817003
  • 2014年03月17日 18:56
  • 2200


一、shared pool的组成           3块区域:free、library cache、row cache           select * from v$sgastat a whe...
  • Jerry_mg
  • Jerry_mg
  • 2017年12月02日 15:31
  • 58
您举报文章:关于软解析与硬解析,以及session cached cursors (翻译,ref asktom) (摘自CNOUG)