The following are performance benefits of direct-path INSERT:

During direct-path INSERT, you can disable the logging of redo and undo entries. Conventional insert operations, in contrast, must always log such entries, because those operations reuse free space and maintain referential integrity.

To create a new table with data from an existing table, you have the choice of creating the new table and then inserting into it, or executing a CREATE TABLE … AS SELECT statement. By creating the table and then using direct-path INSERT operations, you update any indexes defined on the target table during the insert operation. The table resulting from a CREATE TABLE … AS SELECT statement, in contrast, does not have any indexes defined on it; you must define them later.

Direct-path INSERT operations ensure atomicity of the transaction, even when run in parallel mode. Atomicity cannot be guaranteed during parallel direct-path loads (using SQL*Loader).

If errors occur during parallel direct-path loads, some indexes could be marked UNUSABLE at the end of the load. Parallel direct-path INSERT, in contrast, rolls back the statement if errors occur during index update.

Direct-path INSERT must be used if you want to store the data in compressed form using table compression.

Enabling Direct-Path INSERT

You can implement direct-path INSERT operations by using direct-path INSERT statements, inserting data in parallel mode, or by using the Oracle SQL*Loader utility in direct-path mode. Direct-path inserts can be done in either serial or parallel mode.

To activate direct-path INSERT in serial mode, you must specify the APPEND hint in each INSERT statement, either immediately after the INSERT keyword, or immediately after the SELECT keyword in the subquery of the INSERT statement.

When you are inserting in parallel DML mode, direct-path INSERT is the default. In order to run in parallel DML mode, the following requirements must be met:

You must have Oracle Enterprise Edition installed.

You must enable parallel DML in your session. To do this, run the following statement:

ALTER SESSION { ENABLE | FORCE } PARALLEL DML;

You must specify the parallel attribute for the target table, either at create time or subsequently, or you must specify the PARALLEL hint for each insert operation.

To disable direct-path INSERT, specify the NOAPPEND hint in each INSERT statement. Doing so overrides parallel DML mode.

How Direct-Path INSERT Works

You can use direct-path INSERT on both partitioned and non-partitioned tables.

Serial Direct-Path INSERT into Partitioned or Non-partitioned Tables
The single process inserts data beyond the current high water mark of the table segment or of each partition segment. (The high-water mark is the level at which blocks have never been formatted to receive data.) When a COMMIT runs, the high-water mark is updated to the new value, making the data visible to users.

Parallel Direct-Path INSERT into Partitioned Tables
This situation is analogous to serial direct-path INSERT. Each parallel execution server is assigned one or more partitions, with no more than one process working on a single partition. Each parallel execution server inserts data beyond the current high-water mark of its assigned partition segment(s). When a COMMIT runs, the high-water mark of each partition segment is updated to its new value, making the data visible to users.

Parallel Direct-Path INSERT into Non-partitioned Tables
Each parallel execution server allocates a new temporary segment and inserts data into that temporary segment. When a COMMIT runs, the parallel execution coordinator merges the new temporary segments into the primary table segment, where it is visible to users.

Specifying the Logging Mode for Direct-Path INSERT

Direct-path INSERT lets you choose whether to log redo and undo information during the insert operation.

You can specify logging mode for a table, partition, index, or LOB storage at create time (in a CREATE statement) or subsequently (in an ALTER statement).

If you do not specify either LOGGING or NOLOGGING at these times:

The logging attribute of a partition defaults to the logging attribute of its table.

The logging attribute of a table or index defaults to the logging attribute of the tablespace in which it resides.

The logging attribute of LOB storage defaults to LOGGING if you specify CACHE for LOB storage. If you do not specify CACHE, then the logging attributes defaults to that of the tablespace in which the LOB values resides.

You set the logging attribute of a tablespace in a CREATE TABLESPACE or ALTER TABLESPACE statements.

Note:
If the database or tablespace is in FORCE LOGGING mode, then direct path INSERT always logs, regardless of the logging setting.
Direct-Path INSERT with Logging
In this mode, Oracle Database performs full redo logging for instance and media recovery. If the database is in ARCHIVELOG mode, then you can archive redo logs to tape. If the database is in NOARCHIVELOG mode, then you can recover instance crashes but not disk failures.

Direct-Path INSERT without Logging
In this mode, Oracle Database inserts data without redo or undo logging. (Some minimal logging is done to mark new extents invalid, and data dictionary changes are always logged.) This mode improves performance. However, if you subsequently must perform media recovery, the extent invalidation records mark a range of blocks as logically corrupt, because no redo data was logged for them. Therefore, it is important that you back up the data after such an insert operation.

Additional Considerations for Direct-Path INSERT

The following are some additional considerations when using direct-path INSERT.

Index Maintenance with Direct-Path INSERT
Oracle Database performs index maintenance at the end of direct-path INSERT operations on tables (partitioned or non-partitioned) that have indexes. This index maintenance is performed by the parallel execution servers for parallel direct-path INSERT or by the single process for serial direct-path INSERT. You can avoid the performance impact of index maintenance by dropping the index before the INSERT operation and then rebuilding it afterward.

Space Considerations with Direct-Path INSERT
Direct-path INSERT requires more space than conventional-path INSERT.

All serial direct-path INSERT operations, as well as parallel direct-path INSERT into partitioned tables, insert data above the high-water mark of the affected segment. This requires some additional space.

Parallel direct-path INSERT into non-partitioned tables requires even more space, because it creates a temporary segment for each degree of parallelism. If the non-partitioned table is not in a locally managed tablespace in automatic segment-space management mode, you can modify the values of the NEXT and PCTINCREASE storage parameter and MINIMUM EXTENT tablespace parameter to provide sufficient (but not excess) storage for the temporary segments. Choose values for these parameters so that:

The size of each extent is not too small (no less than 1 MB). This setting affects the total number of extents in the object.

The size of each extent is not so large that the parallel INSERT results in wasted space on segments that are larger than necessary.

After the direct-path INSERT operation is complete, you can reset these parameters to settings more appropriate for serial operations.

Locking Considerations with Direct-Path INSERT
During direct-path INSERT, the database obtains exclusive locks on the table (or on all partitions of a partitioned table). As a result, users cannot perform any concurrent insert, update, or delete operations on the table, and concurrent index creation and build operations are not permitted. Concurrent queries, however, are supported, but the query will return only the information before the insert operation.

NOtes Added On
(1) in 10.x and previous releases the APPEND hint gets ignored when it is present in an INSERT with VALUES clause and that is the expected behavior for previous versions, including 9i and 10g. On 11g Release 1, the APPEND hint it is not ignored and a direct-path load is executed for an INSERT with VALUES clause, therefore such statements will allocate one row/database block.

(2) On 11g Release 2, there is a new hint “APPEND_VALUES”, i.e. use direct path load only when the “APPEND_VALUES” hint is referenced. On 11g Release 2 the “APPEND” hint is ignored for INSERT statements with VALUES clause. This is the expected behavior on 11g and it is currently implemented only in 11g Release 2.

最近公司业务的不断扩大，对数据库的使用也不断的增加，导致了基本每天都要进行安装的安装和配置。其实oracle的安装是很简单，不知道是否有更加方便，快捷的安装方式呢？答案：有的，那就是用linux常用的命令tar来进行安装。今天我就来尝试了在linux as4上用tar方式安装oracle 10g，下面简单记录下这次的操作过程。

原数据库所在服务器信息：

kfserver-> more /etc/issue
Red Hat Enterprise Linux AS release 4 (Nahant Update 5)
Kernel \r on an \m

kfserver-> uname -a
Linux kfserver 2.6.9-55.ELsmp #1 SMP Fri Apr 20 17:03:35 EDT 2007 i686 i686 i386 GNU/Linux
新数据库所在服务器信息：

root@testserver:[/root]more /etc/issue
Red Hat Enterprise Linux AS release 4 (Nahant Update 5)
Kernel \r on an \m
root@testserver:[/root]uname -a
Linux testserver 2.6.9-55.ELsmp #1 SMP Fri Apr 20 17:03:35 EDT 2007 i686 i686 i386 GNU/Linux
以上二个系统版本是一样的。

1、在新服务器上添加用户
groupadd oinstall
useradd -g oinstall oracle

2、在新服务器上添加相关目录
mkdir -p /opt/oracle/product

3、设置核心参数
vi /etc/sysctl.conf，添加

# Controls whether core dumps will append the PID to the core filename.
# Useful for debugging multi-threaded applications.
kernel.core_uses_pid = 1
kernel.shmall = 2097152
kernel.shmmax = 2147483648
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 65536
net.ipv4.ip_local_port_range = 1024 65000
net.core.rmem_default = 1048576
net.core.rmem_max = 1048576
net.core.wmem_default = 262144
net.core.wmem_max = 262144

使设置生效：
sysctl -p

4、修改oracle用户的环境变量
su – oracle
vi .bash_profile 添加：

# User specific environment and startup programs
PATH=$PATH:$HOME/bin

export PATH
unset USERNAME
export PS1=”` /bin/hostname -s ` -> ”
export EDITOR=vi
export ORACLE_SID=nuage
export ORACLE_BASE=/opt/oracle
export ORACLE_HOME=$ORACLE_BASE/product/10.2.0/db_1
export ORACLE_CLIENT=$ORACLE_BASE/product/10.2.0/cleint
export LD_LIBRARY=$ORACLE_HOME/lib
export PATH=$ORACLE_HOME/bin:/bin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/X11R6/bin
export NLS_LANG=”SIMPLIFIED CHINESE_CHINA.ZHS16GBK”

5、在原服务器打包程序
把$ORACLE_HOME下的东西打包压缩：
cd $ORACLE_BASE/product
tar cfz oracle10203.tar 10.2

6、传送到新服务器并解压
scp 上传

cd $ORACLE_BASE/product
tar -xzfv oracle10203
7、修改解压后的文件属主

[root@testserver oracle]# chown -R oracle:oinstall /opt/
chown: changing ownership of `/opt/oracle/product/10.2/rdbms/filemap’: No such file or directory

这里出现一个错误，提示文件不存在，经检查这是一个软链接：

[root@testserver oracle]# ls -l /opt/oracle/product/10.2/rdbms/filemap
lrwxrwxrwx 1 root root 22 Jan 10 17:22 /opt/oracle/product/10.2/rdbms/filemap -> /opt/ORCLfmap/prot1_64
经检查，当前的系统确实不存在/opt/ORCLfmap目录，实际上/opt/ORCLfmap是有root.sh生成的，先删除这个软链接：
rm -f /opt/oracle/product/10.2/rdbms/filemap

重新执行root.sh：

[root@testserver opt]# /opt/oracle/product/10.2/root.sh
Running Oracle10 root.sh script…

The following environment variables are set as:
ORACLE_OWNER= oracle
ORACLE_HOME= /opt/oracle/product/10.2

Enter the full pathname of the local bin directory: [/usr/local/bin]:
Copying dbhome to /usr/local/bin …
Copying oraenv to /usr/local/bin …
Copying coraenv to /usr/local/bin …

Creating /etc/oratab file…
Entries will be added to the /etc/oratab file as needed by
Database Configuration Assistant when a database is created
Finished running generic part of root.sh script.
Now product-specific root actions will be performed.
此时软连接和目录都已经存在：

[root@testserver opt]# ls /opt/oracle/product/10.2/rdbms/filemap
bin etc log
[root@testserver opt]# ls -l /opt/oracle/product/10.2/rdbms/filemap
lrwxrwxrwx 1 root root 22 Jan 10 17:40 /opt/oracle/product/10.2/rdbms/filemap -> /opt/ORCLfmap/prot1_64
8、创建数据库
运行dbca，一步一步设置好就可以。

9、清理原网络相关文件、原密码文件等。

一般会出现的问题：
1、chown: changing ownership of `/opt/oracle/product/10.2/rdbms/filemap’: No such file or directory
这个问题在上文已有描述，这里不再赘述。

2、Exception in thread “main” java.lang.UnsatisfiedLinkError: /opt/oracle/product/10.2/jdk/jre/lib/i386/libawt.so: libXp.so.6: cannot open shared object file: No such file or directory

执行dbca时报错：

testserver ->$ dbca
Exception in thread “main”
testserver ->$
这里报错信息非常少，无法判断是什么原因，再执行netca就可以看得更多的错误信息了：

[oracle@testserver bin]$ netca
Exception in thread “main” java.lang.UnsatisfiedLinkError: /opt/oracle/product/10.2/jdk/jre/lib/i386/libawt.so: libXp.so.6: cannot open shared object file: No such file or directory
at java.lang.ClassLoader$NativeLibrary.load(Native Method)
at java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1586)
at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1503)
at java.lang.Runtime.loadLibrary0(Runtime.java:788)
at java.lang.System.loadLibrary(System.java:834)
at sun.security.action.LoadLibraryAction.run(LoadLibraryAction.java:50)
at java.security.AccessController.doPrivileged(Native Method)
at sun.awt.NativeLibLoader.loadLibraries(NativeLibLoader.java:38)
at sun.awt.DebugHelper.(DebugHelper.java:29)
at java.awt.Component.(Component.java:506)

这个问题主要是缺少一个包xorg-x11-deprecated-libs导致的，把包安装上就可以：
[root@testserver soft]# rpm -ivh xorg-x11-deprecated-libs-6.8.2-1.EL.13.37.i386.rpm
warning: xorg-x11-deprecated-libs-6.8.2-1.EL.13.37.i386.rpm: V3 DSA signature: NOKEY, key ID b38a8516
Preparing… ########################################### [100%]
1:xorg-x11-deprecated-lib########################################### [100%]

注意：虽然系统是64位的，但是这里要安装32位的包。

3、ORA-12547: TNS: 丢失连接
在创建数据库前，简单测试一下sqlplus是否正常，发现了这个问题：

testserver ->$ sqlplus “/as sysdba”

SQL*Plus: Release 10.2.0.3.0 – Production on 星期四 1月 10 17:51:35 2008

Copyright (c) 1982, 2006, Oracle. All Rights Reserved.

oraclenewdemo: error while loading shared libraries: libaio.so.1: cannot open shared object file: No such file or directory
ERROR:
ORA-12547: TNS: 丢失连接

请输入用户名:

引起问题的原因是libaio系统包，把相关包安装即可：

[root@testserver soft]# ls
libaio-0.3.105-2.i386.rpm libaio-0.3.105-2.x86_64.rpm libaio-devel-0.3.105-2.x86_64.rpm
[root@testserver soft]# rpm -ivh *
warning: libaio-0.3.105-2.i386.rpm: V3 DSA signature: NOKEY, key ID b38a8516
Preparing… ########################################### [100%]
1:libaio ########################################### [ 33%]
2:libaio ########################################### [ 67%]
3:libaio-devel ########################################### [100%]

同样，这里也需要安装两个32位的包。

4、ORA-01031: insufficient privileges
testserver ->$ sqlplus “/as sysdba”

SQL*Plus: Release 10.2.0.3.0 – Production on 星期五 1月 11 10:48:35 2008

Copyright (c) 1982, 2006, Oracle. All Rights Reserved.

ERROR:
ORA-01031: insufficient privileges

请输入用户名:

这个问题是因为用户组不一致造成的。

原库：

[oracle@kfserver oraInventory]$ id -a
uid=500(oracle) gid=500(oinstall) groups=500(oinstall),501(dba) context=user_u:system_r:unconfined_t
新库：

[oracle@testserver oraInventory]$ id -a
uid=500(oracle) gid=500(oinstall) groups=500(oinstall) context=user_u:system_r:unconfined_t
在新库上为oracle用户添加组就可以了：

[root@testserver ~]# usermod -G dba oracle

create or replace procedure SP_Del_Test
(P_ItemAdmin in mfitem.itemadmin%type, –ItemAdmin
P_ItemCd in mfitem.itemcd%type, –ItemCode
P_Return out number –输出参数
) is
exp exception;
PRAGMA Exception_Init(exp, -2292);
begin
delete from mfitem t
where t.itemcd = P_ItemCd
and t.itemadmin = P_ItemAdmin; –这一句会引发-2292异常，有级连删除异常
EXCEPTION
WHEN EXP THEN
P_Return := 9;
ROLLBACK;
WHEN OTHERS THEN
ROLLBACK;
end SP_Del_Test;　　

使用自定义的异常，自定我们自己的错误消息：过程RAISE_APPLICATION_ERROR

调用RAISE_APPLICATION_ERROR的语法如下：

raise_application_error(error_number, message[, {TRUE | FALSE}]);

error_number是一个范围在-20000至-20999之间的负整数，message是最大长度为2048字节的字符串。如果第三个可选参数为TRUE的话，错误就会被放到前面错误的栈顶。如果为FALSE(默认值)，错误就会替代前面所有的错误。

CREATE PROCEDURE raise_salary (emp_id NUMBER, amount NUMBER) AS
curr_sal NUMBER;
BEGIN
SELECT sal
INTO curr_sal
FROM emp
WHERE empno = emp_id;
IF curr_sal IS NULL THEN
/* Issue user-defined error message. */
raise_application_error (-20101, ‘Salary is missing’);
ELSE
UPDATE emp
SET sal = curr_sal + amount
WHERE empno = emp_id;
END IF;
END raise_salary;

如何抛出PL/SQL异常

DECLARE
out_of_stock EXCEPTION;
number_on_hand NUMBER (4);
BEGIN
…
IF number_on_hand < 1 THEN
RAISE out_of_stock;
END IF;
EXCEPTION
WHEN out_of_stock THEN
– handle the error
END;

1、PL/SQL错误类型

2、异常的声明

有两种异常：用户自定义异常和预定义异常

用户自定义异常就是由程序员自己定义的一个错误。该错误还不是非常重要，所以并没有将整个错误包含在Oracle的错误中。例如，它可能是一个与数据有关的错误。而预定义异常则对应于一般的SQL和PL/SQL错误。

用户自定义异常是在PL/SQL块的声明部分声明的。像变量一样，异常也有一个类型（EXCEPTION）和有效范围。例如： 

DECLARE Exception_name EXCEPTION; …

 3、异常的引发

与异常相关联的错误发生的时候，就会引发相应的异常。用户自定义异常是通过RAISE语句显式引发的，而预定义异常则是在它们关联的ORACLE错误发生的时候隐式引发的。如果发生了一个还没有和异常进行关联的ORACLE错误的时候，也会引发一个异常。该异常可以使用OTHERS子程序进行捕获。预定义的异常也可以使用RAISE进行显式地引发，如果需要这样做的话。

…
RAISE exception_name;
…

 4、异常的处理

发生异常的时候，程序的控制就会转移到代码块的异常处理部分。异常处理部分是由异常处理子程序组成的，这些异常处理子程序可以是针对某些异常的，也可以是针对所有异常的。与该异常相关联的错误发生，并引发了该异常的时候，就会执行异常处理部分的代码。

异常处理部分的语法如下：

 EXCEPTION
WHEN exception_name THEN
Sequence_of_statements1;
WHEN exception_name THEN
Sequence_of_statements2;
WHEN OTHERS THEN
Sequence_of_statements3;
END;

 每一个异常处理部分都是由WHEN子句和引发异常以后要执行的语句组成的。WHEN标识这个处理子程序是针对哪个异常的。

OTHERS异常处理子程序

PL/SQL定义了一个异常处理子程序，即OTHERS。当前异常处理部分定义的所有WHEN语句都没有处理的任意一个已引发的异常，都会导致执行这个OTHERS异常处理子程序。该异常处理子程序应该总是作为代码块的最后一个异常处理子程序，这样就会首先扫描前面的异常处理子程序。WHEN OTHERS会捕获所有异常，不管这些异常是预定义的，还是用户自定义的。

检查错误堆栈—SQLCODE和SQLERRM

PL/SQL使用两个内置函数SQLCODE和SQLERRM提供错误信息。SQLCODE返回的是当前的错误代号，而SQLERRM返回的是当前的错误信息文本。如果是用户自定义的异常，SQLCODE就会返回值1，SQLERRM就会返回“ User-defined Exception”。

 下面是一个使用SQLCODE和SQLERRM的例子

DECLARE
  — Exception to indicate an error condition  

  e_DuplicateAuthors EXCEPTION;

  — IDs for three authors  

  v_Author1 books.author1%TYPE;

  v_Author2 books.author2%TYPE;

  v_Author3 books.author3%TYPE;

  — Code and text of other runtime errors  

  v_ErrorCode log_table.code%TYPE;

  v_ErrorText log_table.message%TYPE;

BEGIN

  /* Find the IDs for the 3 authors of ‘Oracle9i DBA 101′ */

  SELECT author1, author2, author3
 
    INTO v_Author1, v_Author2, v_Author3
 
    FROM books
 
   WHERE title = ‘Oracle9i DBA 101′;

  /* Ensure that there are no duplicates */

  IF (v_Author1 = v_Author2) OR (v_Author1 = v_Author3) OR
   
     (v_Author2 = v_Author3) THEN
 
    RAISE e_DuplicateAuthors;
 
  END IF;

EXCEPTION

  WHEN e_DuplicateAuthors THEN
 
    /* Handler which executes when there are duplicate authors for 
   
    Oracle9i DBA 101.  We will insert a log message recording  
   
    what has happened. */
 
    INSERT INTO log_table
      (info)
   
    VALUES
      (‘Oracle9i DBA 101 has duplicate authors’);
 
  WHEN OTHERS THEN
 
    /* Handler which executes for all other errors. */
 
    v_ErrorCode := SQLCODE;
 
    — Note the use of SUBSTR here.  
 
    v_ErrorText := SUBSTR(SQLERRM, 1, 200);
 
    INSERT INTO log_table
      (code, message, info)
    VALUES
   
      (v_ErrorCode, v_ErrorText, ‘Oracle error occurred’);
 
END;

/

由于该堆栈上每一条错误消息文本的最大长度均为512个字节，但是堆栈中可能会有多条消息文本。在上面的例子中，v_ErrorText只有200个字符。如果该错误消息文本长度大于200个字符，那么赋值语句

 v_ErrorText := SQLERRM;

 就会引发预定义的异常VALUE_ERROR。为了防止发生这种异常，我们使用了内置函数SUBSTR。

注意，SQLCODE和SQLERRM的返回值首先会被分配给局部变量，然后再在SQL语句中使用这些局部变量。因为这些函数都是过程化的函数，所以不能直接在SQL语句中使用它们。

通过下面这个例子我们看看错误号和相应的错误消息文本之间的关系

 
set serveroutput on  
 
BEGIN  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM(0): ‘ || SQLERRM(0));  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM(100): ‘ || SQLERRM(100));  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM(10): ‘ || SQLERRM(10));  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM: ‘ || SQLERRM);  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM(-1): ‘ || SQLERRM(-1));  
 
  DBMS_OUTPUT.PUT_LINE(‘SQLERRM(-54): ‘ || SQLERRM(-54));  
 
END;  
 
/  
 
–运行结果如下  
 
SQL> @SQLERRM.sql  
 
SQLERRM(0): ORA-0000: normal, successful completion  
 
SQLERRM(100): ORA-01403: no data found  
 
SQLERRM(10):  -10: non-ORACLE exception  
 
SQLERRM: ORA-0000: normal, successful completion  
 
SQLERRM(-1): ORA-00001: unique constraint (.) violated  
 
SQLERRM(-54): ORA-00054: resource busy and acquire with NOWAIT specified  
 
   
 
PL/SQL procedure successfully completed.  
 

EXCEPTION_INIT pragma

你可以将一个经过命名的异常和一个特别的ORACLE错误相关联。这会使你专门能够捕获此错误，而不是通过WHEN OTHERS处理器来进行捕获。EXCEPTION_INIT pragma的语法如下：

PRAGMA       EXCEPTION_INIT(exception_name, -Oracle_error_number);

这里，exception_name是在PRAGMA前面声明的异常的名字，而Oracle_error_number是与此命名异常相关的所需错误代码。这个PRAGMA必须在声明部分。

下面这个例子在运行时刻如果遇到“ORA-1400:mandatory NOT NULL column missing or NULL during insert”错误时将引发e_MissingNull–用户定义的异常。

DECLARE  
 
  e_MissingNull EXCEPTION;  
 
  PRAGMA EXCEPTION_INIT(e_MissingNull, -1400);  
 
BEGIN  
 
  INSERT INTO students (id) VALUES (NULL);  
 
EXCEPTION   s
 
  WHEN e_MissingNull then  
 
    INSERT INTO log_table (info) VALUES (‘ORA-1400 occurred’);  
 
END;  
 
/ 
 
 

每次发生PRAGMA EXCEPTION_INIT时，一个Oracle错误只能和一个用户自定义异常相关联。在异常处理内部，SQLCODE和SQLERRM将会返回发生Oracle错误的代码和错误消息，但是不会返回用户定义的消息。

使用RAISE_APPLICATION_ERROR

你可以使用内置函数RAISE_APPLICATION_ERROR以创建自己的错误消息，这可能要比已命名的异常更具说明性。用户定义消息从块中传递到调用环境中的方式和ORACLE错误是一样的。语法如下：

RAISE_APPLICATION_ERROR(error_number,error_message,[keep_errors]);

error_number是从-200000到-20999之间的参数，error_message是与此错误相关的正文，不能多于512个字节。而keep_errors是一个布尔值，是可选的，如果为TRUE，那么新的错误将被添加到已经引发的错误列表中（如果有的话）。如果为FALSE（这是缺省的设置），那么新的错误将替换错误的当前列表。

例如下面的这个例子将在为一个新的学生注册以前检查是否在班级中有足够的地方容纳他。

CREATE OR REPLACE PROCEDURE Register(
                                    
                                     /* Registers the student identified by the p_StudentID parameter in the class 
                                     
                                          identified by the p_Department and p_Course parameters. Before calling 
                                     
                                          ClassPackage.AddStudent, which actually adds the student to the class, this 
                                     
                                          procedure verifies that there is room in the class, and that the class 
                                     
                                          exists. */
                                    
                                     p_StudentID IN students.id%TYPE,
                                    
                                     p_Department IN classes.department%TYPE,
                                    
                                     p_Course IN classes.course%TYPE) AS

  v_CurrentStudents NUMBER; — Current number of students in the class 

  v_MaxStudents NUMBER; — Maximum number of students in the class 

BEGIN

  /* Determine the current number of students registered, and the maximum 
 
  number of students allowed to register. */

  SELECT current_students, max_students
 
    INTO v_CurrentStudents, v_MaxStudents
 
    FROM classes
 
   WHERE course = p_Course
       
     AND department = p_Department;

  /* Make sure there is enough room for this additional student. */

  IF v_CurrentStudents + 1 > v_MaxStudents THEN
 
    RAISE_APPLICATION_ERROR(-20000,
                            ‘Can”t add more students to ‘ ||
                           
                            p_Department || ‘ ‘ || p_Course);
 
  END IF;

  /* Add the student to the class. */

  ClassPackage.AddStudent(p_StudentID, p_Department, p_Course);

EXCEPTION

  WHEN NO_DATA_FOUND THEN
 
    /* Class information passed to this procedure doesn’t exist. Raise an error 
   
    to let the calling program know of this. */
 
    RAISE_APPLICATION_ERROR(-20001,
                            p_Department || ‘ ‘ || p_Course ||
                           
                            ‘ doesn”t exist!’);
 
END Register;

/

5、异常的传播

1）在执行部分引发的异常

当一个异常是在块的执行部分引发的，PL/SQL使用下面的方法决定要激活哪个异常处理器：

如果当前块对该异常设置了处理器，那么执行它并成功完成该块的执行，然后控制会转给包含块。

如果当前块没有对当前异常定义处理器，那么通过在包含块中引发它来传播异常。然后对包含块执行步骤一。

2）在声明部分引发的异常

如果在声明部分的赋值操作引发了一个异常，那么该异常将立即传播给包含块。发生这种情况以后，在前面给出的法则将进一步被应用到异常的传播上。尽管在当前块中有一个处理器，它也不会被执行。

3）在异常处理部分引发的异常

在异常处理器中也可能引发异常，这可以是通过RAISE语句显式引发的，也可以是由运行时刻错误隐含引发的。无论怎样，异常都立即被传播给包含块，这和声明部分引发的异常相类似。

6、使用异常的准则

1）异常的范围

异常像变量一样，也是有一定范围的。如果用户自定义异常传播到它的范围之外，就不能再通过名称引用它。

BEGIN
  DECLARE
 
    e_UserDefinedException EXCEPTION;
 
  BEGIN
 
    RAISE e_UserDefinedException;
 
  END;

EXCEPTION

  /* e_UserDefinedException is out of scope here – can only be 
 
  handled by an OTHERS handler */

  WHEN OTHERS THEN
 
    /* Just re-raise the exception, which will be propagated to the 
   
    calling environment */
 
    RAISE;
 
END;

/

一般而言，如果打算将用户自定义的错误传播到代码块之外，最好的方法就是在包中定义该异常，以使其在该代码块之外仍可见，或使用RAISE_APPLICATION_ERROR引发该异常。如果创建一个成为GLOBALS的包，并在其中定义了一个e_UserDefinedException异常，那么这个异常在外部块中仍然可见。

如下例所示

  
/* This package contains global declarations. Objects declared here will  
 
   be visible via qualified references for any other blocks or procedures. 
 
   Note that this package does not have a package body. */ 
 
   
 
  /* A user-defined exception. */ 
 
  e_UserDefinedException EXCEPTION;  
 
END Globals;  
 
/  
 
   
 
–有了这个和GLOBALS包以后，就可以重写前面的代码：  
 
   
 
BEGIN  
 
  BEGIN  
 
    RAISE Globals.e_UserDefinedException;  
 
  END;  
 
EXCEPTION  
 
  /* Since e_UserDefinedException is still visible, we can handle it  
 
     explicitly */ 
 
  WHEN Globals.e_UserDefinedException THEN           –引用包中定义异常  
 
    /* Just re-raise the exception, which will be propagated to the 
 
       calling environment */ 
 
    RAISE;  
 
END;  
 
/ 
 
 

2）避免未处理的异常

优秀的编程经验是在整个程序中避免出现任何未经过处理的异常。这可以通过在程序的最顶层使用一个OTHERS子程序来实现。该处理子程序可以只登记错误并登记错误发生的位置，通过这种方法，就可以保证每个错误都会得到检查。

如下例所示

CREATE OR REPLACE PACKAGE Globals AS

  v_errortext varchar2(200);

Begin

  …
 
  EXCEPTION
 
  WHEN OTHERS THEN
 
  v_errornumber := SQLCODE;

  v_errortext := SUBSTR(SQLERRM, 1, 200);

  INSERT INTO log_table
    (code, message, info)
 
  VALUES
 
    (v_errornumber,
     v_errortext,
     ’Oracle error occurred
     at’ || TO_CHAR(SYSDATE, ’DD – MON – YY HH24 :MI :SS’));

END;

3）标识错误发生的位置

由于整个代码块都使用同一个异常处理部分检查并处理异常，所以很难确定引发这个错误的是哪一条SQL语句。考虑下面示例

 
BEGIN
SELECT…  
 
SELECT..  
 
SELECT…  
 
EXCEPTION  
 
WHEN NO_DATA_FOUND THEN  
 
–which select statement raised the exception?  
 
END;  
 
   
 
–解决上述问题的方法有两种。第一种是添加一个标识该SQL语句的计数器：  
 
   
 
DECLARE  
 
V_selectcounter NUMBER:=1;  
 
BEGIN  
 
SELECT…  
 
V_selectcounter NUMBER:=2;  
 
SELECT…  
 
V_selectcounter NUMBER:=3;  
 
SELECT…  
 
EXCEPTION  
 
WHEN NO_DATA_FOUND THEN  
 
INSERT INTO log_table(info) VALUES(‘NO DATA FOUND IN SELECT’||v_selectcounter);  
 
END;  
 
   
 
–另一种方法是将每一条语句都放置在它自己的子块中：  
 
   
 
BEGIN  
 
BEGIN  
 
              SELECT…  
 
       EXCEPTION  
 
              WHEN NO_DATA_FOUND THEN  
 
                     INSERT INTO log_table(info) VALUES(‘NO DATA FOUND IN SELECT 1’);  
 
       END;  
 
BEGIN  
 
              SELECT…  
 
       EXCEPTION  
 
              WHEN NO_DATA_FOUND THEN  
 
                     INSERT INTO log_table(info) VALUES(‘NO DATA FOUND IN SELECT 2’);  
 
       END;  
 
BEGIN  
 
              SELECT…  
 
       EXCEPTION  
 
              WHEN NO_DATA_FOUND THEN  
 
                     INSERT INTO log_table(info) VALUES(‘NO DATA FOUND IN SELECT 3’);  
 
       END;  
 
END;  
  
  

7、异常代码的编写风格

1）RAISE_APPLICATION_ERROR和RAISE的比较

通常而言，推荐对设计给终端用户看的错误，使用RAISE_APPLICATION_ERROR。因为对于他们而言，具体的错误编号和描述性文本非常有用。而另一方面，对设计为由程序直接进行处理的异常，推荐使用RAISE。

2）将异常作为控制语句

因为引发异常会将程序的控制逻辑转移到代码块的异常处理部分，所以可以将RAISE语句用作控制语句，就像GOTO语句一样。例如，如果我们有很深的嵌套循环，并需要立即从中退出的时候，这可能会非常有用。

格式化函数提供一套有效的工具用于把各种数据类型（日期/时间，int，float，numeric）转换成格式化的字符串以及反过来从格式化的字符串转换成原始的数据类型。

表 5-6. 格式化函数

所有格式化函数都是 format-picture （格式图）的第二个参数。

表 5-7. 用于日期/时间 to_char() 版本的format-pictures。

所有 format-pictures （格式图）允许使用后缀（后缀/前缀）。对于近似
format-picture（格式图），后缀总是有效的。’fx’只是全局前缀。

表 5-8. 用于日期/时间 to_char() 版本的 format-pictures （格式图）后缀。

‘/’ – 必须用做双 //，例如 ‘//hh//mi//ss’

‘”‘ – 双引号之间的字串被忽略并且不被分析。如果你想向输出写 ‘”‘ 你必须用 //”，例如 ‘//”yyyy
month//”‘。

text – postgresql 的 to_char() 支持不带 ‘”‘
的文本，但是带有双引号的字串会快些并且可以保证该文本不会被解释成关键字（format-picture，格式图），例如
‘”hello year: “yyyy’。

表 5-9. 用于数字 (int/float/numeric) to_char()
版本的format-pictures （格式图）。

注意：通过a sign formatted via ‘sg’，’pl’ 或 ‘mi’
格式化的符号数不一定是数字；to_char(-12, ‘s9999′)生成：

' -12'

，但是 to_char(-12, ‘mi9999′)生成：

'- 12'

。oracle 不允许在 ’9′ 前面使用 ‘mi’，在 oracle 里，它总是在 ’9′ 后面。.

表 5-10. to_char()的例子

'tuesday , 05:39:18'

'tuesday, 05:39:18'

' -.10'

'-.1'

' 0.1'

' 0012.0'

'0012'

' 485'

'-485'

' 4 8 5'

' 1,485'

' 1 485'

' 148.500'

' 148,500'

' 3 148,500'

'485-'

'485-'

'485'

'+485'

'+485'

'-485'

'4-85'

'<485>'

'dm 485'

' cdlxxxv'

'cdlxxxv'

'v'

' 482nd'

'good number: 485'

'pre-decimal: 485 post-decimal: .800'

' 12000'

' 12400'

' 125'

还原点是10GR2新增加的一个功能,它是建立在数据库层面上的, 跟savepoint不同. 必须在数据库处于ARCHIVELOG模式下才可以使用,主题思想是在数据库的某个时间点创建一个restore point,在随后的某个时间点可以将表或数据库falshback 到这个restore point.那么能否真正的完成flashback,依赖于创建的restore point类型和flashback要操作的对象.flashback database需要SYSDBA权限,flashback table需要 FLASHBACK ANY TABLE的权限.

guarantee restore point for flashback database

guarantee restore point必须在数据库处于flashback on 的状态下，才可以定义的一中restore point,它的主题思想是保证数据库能够准确完整的flashback 到定义的 restore point. 操作依赖于flashback log和archivelog, 如果定义了guarantee restore point,那么flashback logs将不受db_recovery_retention_target参数的限制,oracle不会删除flashback log,只要flashback arae有足够的磁盘空间,oracle就会保存足够的flashback logs,满足flashback database to restore point.

如果数据库没有处于flashback on,那么定义一个guarantee restore point必须在database mount状态下进行.其实这个也可以理解,oracle根据定义的restore point,可以个保证database open后的归档可以不被删除.但有一点必须明确的是,guarantee restore point是针对flashback database而言的,不对flashback table起作用，为什么呢, 因为flashback table依赖于undo,如果在guarantee时间范围内,undo过期了, 那么flatable table 还是不能还原到定义的guarantee restore point.

Normal restore point

普通的restore point 非常简单, 如果要flashback database,那么数据库必须处于flashback on 状态,但这个时候不一定可以flashback database到定义的restore point,oracle会根据db_recovery_retention_target自动维护创建的restore ponit和flashback logs,如果超过db_recovery_retention_target 定义的restore point,oracle会自动删除 , 无论是flashback table或者flashback database都不能完成.

简单的流水帐记录,为了试自己更加明确restore point的原理.

创建一个guarantee restore point

SQL> create restore point a guarantee flashback database;
创建一个普通的restore point

SQL> create restore point b;
flashback dtaase to restore point:(必须在guarantee restore point for flashback database的情况)

SQL> startup mount
ORACLE instance started.

Total System Global Area 209715200 bytes

Fixed Size 1266632 bytes

Variable Size 134220856 bytes

Database Buffers 71303168 bytes

Redo Buffers 2924544 bytes

Database mounted.

SQL> flashback database to restore point a;

Flashback complete.

SQL> alter database open resetlogs;

Database altered.

falshback table

SQL> alter table test_histogram enable row movement;

Table altered.

SQL> flashback table test_histogram to restore point a;

Flashback complete.

 SYNOPSIS

      vmstat [-dnS] [interval [count]]

      vmstat -f | -s | -z

 DESCRIPTION

      The vmstat command reports certain statistics kept about process,

      virtual memory, trap, and CPU activity.  It also can clear the

      accumulators in the kernel sum structure.