重建控制文件

首先我们应该先了解一下控制文件的重要性

控制文件的概念：控制文件是一个很小的二进制文件，用于记录数据库的物理结构。一个控制文件只属于一个数据库。创建数据库时，创建控制文件。当数据库的物理结构改变的时候，Oracle会更新控制文件。用户不能编辑控制文件，控制文件的修改由Oracle完成。

数据库的启动和正常运行都离不开控制文件。启动数据库时，Oracle从初始化参数文件中获得控制文件的名字及位置，打开控制文件，然后从控制文件中读取数据文件和联机日志文件的信息，最后打开数据库。数据库运行时，Oracle会修改控制文件，所以，一旦控制文件损坏，数据库将不能正常运行。如果没有数据库的备份和归档日志文件，数据库将无法恢复。因此，我们应该多路镜像控制文件（Multiplex Control Files），并把每个镜像的控制文件分布在不同的物理磁盘。根据经验，控制文件多路镜像以后，几个控制文件同时坏掉的可能性几乎为零。控制文件管理的重心是重在预防，而不是亡羊补牢！

控制文件的内容：

数据库名称（Database Name）

创建数据库的时间戳

数据文件的名字及位置

联机日志文件的名字及位置

表空间信息

日志历史记录（Log History）

归档日志的信息

备份信息

当前的日志序列号（Log Sequence Number）

检查点信息（Checkpoint）

控制文件的大小由下面的几个参数决定：

MAXLOGFILES

MAXLOGMEMBERS

MAXLOGHISTORY

MAXDATAFILES

MAXINSTANCES

当增加、重命名、删除一个数据文件或者一个联机日志文件时，Oracle服务器进程（Server Process）会立即更新控制文件以反映数据库结构的这种变化。所以，Oracle总是告诫我们，在数据库的结构发生变化后，要备份控制文件。日志写进程LGWR负责把当前日志序列号记录到控制文件中。校验点进程CKPT负责把校验点的信息记录到控制文件中。归档进程负责把归档日志的信息记录到控制文件中。

初始化参数CONTROL_FILES的值记录控制文件的位置。通常，DBA应该镜像控制文件，把每个控制文件分布到不同的物理磁盘，发生灾难时，即使其中一个控制文件损坏，数据不会丢失，也不会使整个数据库陷于瘫痪。

 

我们了解了控制文件的重要性后，如果我们真的将控制文件丢失，我们该如何做呢？

1、  从备份中拷贝控制文件

2、  重建控制文件

一、重建控制文件：

Oracle提供如下命令:

alter databasebackup controlfile to trace;

操作步骤如下：

1、        连接数据库并查看版本

[oracle@edbjr2p1~]$ sqlplus / as sysdba

 

SQL*Plus:Release 10.2.0.1.0 - Production on Sat Mar 15 18:45:48 2014

 

Copyright (c)1982, 2005, Oracle.  All rights reserved.

 

 

Connected to:

Oracle Database10g Enterprise Edition Release 10.2.0.1.0 - Production

With thePartitioning, OLAP and Data Mining options

 

SYS@PROD>selectversion from v$instance;

 

VERSION

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

10.2.0.1.0

 

SYS@PROD>archivelog list

Database logmode              Archive Mode

Automaticarchival             Enabled

Archivedestination            /u01/app/oracle/product/10.2.0/db_1/dbs/arch

Oldest onlinelog sequence     2

Next logsequence to archive   4

Current logsequence           4

SYS@PROD>

2、得到创建控制文件的语句与操作

SYS@PROD>alterdatabase backup controlfile to trace;

 

Databasealtered.

3、查看跟踪文件的内容

[oracle@edbjr2p1udump]$ more prod_ora_18466.trc

Dump file/u01/app/oracle/admin/PROD/udump/prod_ora_17286.trc

Oracle Database10g Enterprise Edition Release 10.2.0.1.0 - Production

With thePartitioning, OLAP and Data Mining options

ORACLE_HOME =/u01/app/oracle/product/10.2.0/db_1

Systemname:    Linux

Node name:      edbjr2p1

Release:        2.6.18-238.el5

Version:        #1 SMP Tue Jan 4 15:24:05 EST 2011

Machine:        i686

Instance name:PROD

Redo threadmounted by this instance: 1

Oracle processnumber: 26

Unix processpid: 17286, image: oracle@edbjr2p1 (TNS V1-V3)

 

*** 2014-03-1520:15:21.711

*** SERVICENAME:(SYS$USERS) 2014-03-15 20:15:21.711

*** SESSIONID:(289.3) 2014-03-15 20:15:21.711

tkcrrsarc:(WARN) Failed to find ARCH for message (message:0x1)

tkcrrpa: (WARN)Failed initial attempt to send ARCH message (message:0x1)

*** 2014-03-1520:15:34.425

-- The followingare current System-scope REDO Log Archival related

-- parametersand can be included in the database initialization file.

--

--LOG_ARCHIVE_DEST=''

--LOG_ARCHIVE_DUPLEX_DEST=''

--

--LOG_ARCHIVE_FORMAT=%t_%s_%r.dbf

--

--DB_UNIQUE_NAME="PROD"

--

--LOG_ARCHIVE_CONFIG='SEND, RECEIVE, NODG_CONFIG'

--LOG_ARCHIVE_MAX_PROCESSES=2

--STANDBY_FILE_MANAGEMENT=MANUAL

-- STANDBY_ARCHIVE_DEST=?/dbs/arch

-- FAL_CLIENT=''

-- FAL_SERVER=''

--

--LOG_ARCHIVE_DEST_1='LOCATION=/u01/app/oracle/product/10.2.0/db_1/dbs/arch'

--LOG_ARCHIVE_DEST_1='MANDATORY NOREOPEN NODELAY'

--LOG_ARCHIVE_DEST_1='ARCH NOAFFIRM EXPEDITE NOVERIFY SYNC'

-- LOG_ARCHIVE_DEST_1='NOREGISTERNOALTERNATE NODEPENDENCY'

--LOG_ARCHIVE_DEST_1='NOMAX_FAILURE NOQUOTA_SIZE NOQUOTA_USED NODB_UNIQUE_NAME'

--LOG_ARCHIVE_DEST_1='VALID_FOR=(PRIMARY_ROLE,ONLINE_LOGFILES)'

--LOG_ARCHIVE_DEST_STATE_1=ENABLE

--

-- Below are twosets of SQL statements, each of which creates a new

-- control fileand uses it to open the database. The first set opens

-- the databasewith the NORESETLOGS option and should be used only if

-- the currentversions of all online logs are available. The second

-- set opens thedatabase with the RESETLOGS option and should be used

-- if onlinelogs are unavailable.

-- Theappropriate set of statements can be copied from the trace into

-- a scriptfile, edited as necessary, and executed when there is a

-- need tore-create the control file.

--

--     Set #1. NORESETLOGS case

--

-- The followingcommands will create a new control file and use it

-- to open thedatabase.

-- Data used byRecovery Manager will be lost.

-- Additionallogs may be required for media recovery of offline

-- Use this onlyif the current versions of all online logs are

-- available.

-- Aftermounting the created controlfile, the following SQL

-- statementwill place the database in the appropriate

-- protectionmode:

--  ALTER DATABASE SET STANDBY DATABASE TOMAXIMIZE PERFORMANCE

STARTUP NOMOUNT

CREATECONTROLFILE REUSE DATABASE "PROD" NORESETLOGS  ARCHIVELOG

    MAXLOGFILES 30

    MAXLOGMEMBERS 5

    MAXDATAFILES 100

    MAXINSTANCES 1

    MAXLOGHISTORY 292

LOGFILE

  GROUP 1 (

   '/u01/app/oracle/oradata/PROD/disk1/redo01.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo01_b.log'

  ) SIZE 100M,

  GROUP 2 (

   '/u01/app/oracle/oradata/PROD/disk1/redo02.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo02_b.log'

  ) SIZE 100M,

  GROUP 3 (

   '/u01/app/oracle/oradata/PROD/disk1/redo03.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo03_b.log'

  ) SIZE 100M,

  GROUP 4 (

   '/u01/app/oracle/oradata/PROD/disk1/redo04_a.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo04_b.log'

  ) SIZE100M

-- STANDBYLOGFILE

DATAFILE

 '/u01/app/oracle/oradata/PROD/disk1/system01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/undotbs01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/sysaux01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/EXAMPLE_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk3/USERS_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk5/dds_01.dbf'

CHARACTER SETUS7ASCII

;

-- Commands tore-create incarnation table

-- Below lognames MUST be changed to existing filenames on

-- disk. Any onelog file from each branch can be used to

-- re-createincarnation records.

-- ALTERDATABASE REGISTER LOGFILE'/u01/app/oracle/product/10.2.0/db_1/dbs/arch1_1_842277968.dbf';

-- Recovery isrequired if any of the datafiles are restored backups,

-- or if thelast shutdown was not normal or immediate.

RECOVER DATABASE

-- All logs needarchiving and a log switch is needed.

ALTER SYSTEMARCHIVE LOG ALL;

-- Database cannow be opened normally.

ALTER DATABASEOPEN;

-- Commands toadd tempfiles to temporary tablespaces.

-- Onlinetempfiles have complete space information.

-- Othertempfiles may require adjustment.

ALTER TABLESPACETEMP01 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp01.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP1 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp1_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP2 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp2_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

-- End oftempfile additions.

--

*** 2014-03-1520:18:32.065

*** 2014-03-1520:18:32.065 60679 kcrr.c

ARCH: Archivaldisabled due to shutdown: 1089

*** 2014-03-1520:18:32.082 60679 kcrr.c

ARCH: Archivaldisabled due to shutdown: 1089

4、编辑这个trace文件,我们就可以获得创建控制文件的脚本.

根据数据库不同状况，你可以选择是使用RESETLOGS/NORESETLOGS来重建控制文件

我们获得以下脚本:

 

[oracle@standby tools]$ cat createctlf.sql

STARTUP NOMOUNT

CREATECONTROLFILE REUSE DATABASE "PROD" NORESETLOGS  ARCHIVELOG

    MAXLOGFILES 30

    MAXLOGMEMBERS 4

    MAXDATAFILES 100

    MAXINSTANCES 1

    MAXLOGHISTORY 292

LOGFILE

  GROUP 1 (

   '/u01/app/oracle/oradata/PROD/disk1/redo01.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo01_b.log'

  ) SIZE 100M,

  GROUP 2 (

   '/u01/app/oracle/oradata/PROD/disk1/redo02.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo02_b.log'

  ) SIZE 100M,

  GROUP 3 (

   '/u01/app/oracle/oradata/PROD/disk1/redo03.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo03_b.log'

  ) SIZE 100M,

  GROUP 4 (

   '/u01/app/oracle/oradata/PROD/disk1/redo04_a.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo04_b.log'

  ) SIZE 100M

-- STANDBYLOGFILE

DATAFILE

 '/u01/app/oracle/oradata/PROD/disk1/system01.dbf',

  '/u01/app/oracle/oradata/PROD/disk1/undotbs01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/sysaux01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/EXAMPLE_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk3/USERS_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk5/dds_01.dbf'

CHARACTER SET US7ASCII;

RECOVERDATABASE;

ALTER SYSTEMARCHIVE LOG ALL;

ALTER DATABASEOPEN;

ALTER TABLESPACETEMP01 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp01.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP1 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp1_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP2 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp2_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

5、运行此脚本即可重建控制文件:

SYS@PROD>startupnomount;

ORACLE instancestarted.

 

Total SystemGlobal Area  419430400 bytes

Fixed Size                  1219784 bytes

VariableSize             121635640 bytes

DatabaseBuffers          293601280 bytes

RedoBuffers                2973696 bytes

SYS@PROD>CREATECONTROLFILE REUSE DATABASE "PROD" NORESETLOGS  ARCHIVELOG

    MAXLOGFILES 30

    MAXLOGMEMBERS 4

    MAXDATAFILES 100

    MAXINSTANCES 1

    MAXLOGHISTORY 292

LOGFILE

  GROUP 1 (

    '/u01/app/oracle/oradata/PROD/disk1/redo01.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo01_b.log'

  ) SIZE 100M,

  GROUP 2 (

   '/u01/app/oracle/oradata/PROD/disk1/redo02.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo02_b.log'

  ) SIZE 100M,

  GROUP 3 (

   '/u01/app/oracle/oradata/PROD/disk1/redo03.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo03_b.log'

  ) SIZE 100M,

  GROUP 4 (

   '/u01/app/oracle/oradata/PROD/disk1/redo04_a.log',

   '/u01/app/oracle/oradata/PROD/disk2/redo04_b.log'

  ) SIZE 100M

-- STANDBYLOGFILE

DATAFILE

 '/u01/app/oracle/oradata/PROD/disk1/system01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/undotbs01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/sysaux01.dbf',

 '/u01/app/oracle/oradata/PROD/disk1/EXAMPLE_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk3/USERS_01.dbf',

 '/u01/app/oracle/oradata/PROD/disk5/dds_01.dbf'

CHARACTER SETUS7ASCII;

  2   3    4    5   6    7    8   9   10   11  12   13   14  15   16   17  18   19   20  21   22   23  24   25   26  27   28   29  30   31   32 

Control filecreated.

 

SYS@PROD>RECOVERDATABASE;

ORA-00283:recovery session canceled due to errors

ORA-00264: norecovery required

 

 

SYS@PROD>ALTERSYSTEM ARCHIVE LOG ALL;

 

System altered.

 

SYS@PROD>ALTERDATABASE OPEN;

 

Databasealtered.

 

SYS@PROD>ALTERTABLESPACE TEMP01 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp01.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP1 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp1_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

ALTER TABLESPACETEMP2 ADD TEMPFILE '/u01/app/oracle/oradata/PROD/disk1/temp2_01_grp.dbf'

     SIZE 20971520  REUSE AUTOEXTEND ON NEXT 8192  MAXSIZE 32767M;

  2 

Tablespacealtered.

 

SYS@PROD>  2 

Tablespacealtered.

 

SYS@PROD>  2 

Tablespacealtered.

 

SYS@PROD>

 

二、备份控制文件

以上给出生成创建控制文件脚本并重建控制文件的方法，但是具体恢复中遇到的问题可能需要具体对待,当得不到trace文件，可根据这个格式，查看数据文件，更改相关内容再重建.这种方法通常是在没有控制文件(二进制文件)备份的情况下所采用的，如果存在备份应该使用备份的控制文件尝试恢复.

注：

SQL>alterdatabase backup controlfile to '\home\oracle\backup\control.bkp'; //备份二进制控制文件