Welcome 微信登录

首页 / 数据库 / MySQL / Oracle 单实例 从32位 迁移到 64位 方法

关于单实例从32位迁移到64位的说明，参考：Oracle Convert a 32-bit Database to 64-bit Database（32位 转到 64位）说明 http://www.linuxidc.com/Linux/2012-03/55757.htm在这篇文章中演示第一种方法：直接copy datafile进行迁移。一.测试环境：

Source：

       DB：11.2.0.3 32 位       OS:  Oracle Linux 6.1 32 位       IP:192.168.3.200       ORACLE_HOME：[oracle@tianlesoftware ~]$ echo$ORACLE_HOME/u01/app/oracle/product/11.2.0/db_1 

Target ：

       DB:11.2.0.3 64位       OS：Oracle Linux 6.1 64位       IP:192.168.3.201ORACLE_HOME:rac1:/home/oracle> echo $ORACLE_HOME/u02/app/oracle/product/11.2.0/db_1这里2个ORACLE_HOME 不一样，所以我们在操作过程中需要重建控制文件。Oracle 控制文件 http://www.linuxidc.com/Linux/2011-03/33069.htm

二．具体操作如下：

1. 冷备份整个DB，以防修改wordsize失败。2. 将source 端的控制文件dump 到 trace文件里。 SQL> oradebug setmypidStatement processed.SQL> alter database backup controlfile to trace;Database altered.SQL> oradebug tracefile_name/u01/app/oracle/diag/rdbms/anqing/anqing/trace/anqing_ora_8874.trc查看trace 文件，获取控制文件重建脚本：[oracle@tianlesoftware u01]$ cat /u01/app/oracle/diag/rdbms/anqing/anqing/trace/anqing_ora_8874.trcSTARTUP NOMOUNTCREATE CONTROLFILE REUSE DATABASE"ANQING" NORESETLOGS ARCHIVELOG   MAXLOGFILES 16   MAXLOGMEMBERS 3   MAXDATAFILES 100   MAXINSTANCES 8   MAXLOGHISTORY 292LOGFILE GROUP 1 "/u01/app/oracle/oradata/anqing/redo01.log"  SIZE 50M BLOCKSIZE 512,  GROUP 2"/u01/app/oracle/oradata/anqing/redo02.log" SIZE 50M BLOCKSIZE 512, GROUP 3 "/u01/app/oracle/oradata/anqing/redo03.log"  SIZE 50M BLOCKSIZE 512-- STANDBY LOGFILEDATAFILE "/u01/app/oracle/oradata/anqing/system01.dbf", "/u01/app/oracle/oradata/anqing/sysaux01.dbf", "/u01/app/oracle/oradata/anqing/undotbs01.dbf", "/u01/app/oracle/oradata/anqing/users01.dbf"CHARACTER SET ZHS16GBK;RECOVER DATABASEALTER SYSTEM ARCHIVE LOG ALL;ALTER DATABASE OPEN;ALTER TABLESPACE TEMP ADD TEMPFILE "/u01/app/oracle/oradata/anqing/temp01.dbf"SIZE 30408704  REUSE AUTOEXTEND ONNEXT 655360  MAXSIZE 32767M;3.    在Target 端安装64位的Oracle软件，这里仅安装，不创建实例。64位 linux平台下Oracle安装文档 http://www.linuxidc.com/Linux/2012-03/55758.htm4.   Clean shutdown Source 端的32位DBOracle 实例恢复时 前滚（roll forward） 后滚（rollback） 问题 http://www.linuxidc.com/Linux/2011-03/33907.htm SQL> shutdown immediateDatabase closed.Database dismounted.ORACLE instance shut down.5.     将Source 端的所有datafile copy 到Target 端的对应目录。注意这里的TEMP 表空间，虽然我们可以一起拷贝过去，但是还是建议重建TEMP，已避免其他的问题。5.1  在Target 端创建相关的目录 rac1:/u02> cd app/oracle/rac1:/u02/app/oracle> lscheckpoints diag  productrac1:/u02/app/oracle> mkdir adminrac1:/u02/app/oracle> mkdir fast_recovery_area rac1:/u02/app/oracle/product/11.2.0/db_1/dbs>cd /u02/app/oracle/fast_recovery_area/rac1:/u02/app/oracle/fast_recovery_area>mkdir anqing rac1:/u02/app/oracle> mkdir oradatarac1:/u02/app/oracle> lsadmin checkpoints  diag  fast_recovery_area  oradata product rac1:/u02/app/oracle> cd oradatarac1:/u02/app/oracle/oradata> mkdiranqing rac1:/u02/app/oracle/oradata> cd ../rac1:/u02/app/oracle> lsadmin checkpoints  diag  fast_recovery_area  oradata productrac1:/u02/app/oracle> cd adminrac1:/u02/app/oracle/admin> lsrac1:/u02/app/oracle/admin> mkdir anqingrac1:/u02/app/oracle/admin> cd anqingrac1:/u02/app/oracle/admin/anqing> mkdiradumprac1:/u02/app/oracle/admin/anqing> mkdirdpdumprac1:/u02/app/oracle/admin/anqing> mkdirpfilerac1:/u02/app/oracle/admin/anqing> mkdirscriptsrac1:/u02/app/oracle/admin/anqing> lsadump dpdump  pfile  scripts5.2  copy 所有datafiles[oracle@tianlesoftware anqing]$ scp *192.168.3.201:/u02/app/oracle/oradata/anqingoracle@192.168.3.201"s password: control01.ctl                                                                            100% 9520KB   9.3MB/s   00:01   redo01.log                                                                               100%   50MB   5.6MB/s  00:09    redo02.log                                                                               100%   50MB   6.3MB/s  00:08    redo03.log                                                                               100%   50MB   5.0MB/s  00:10    sysaux01.dbf                                                                              100%  510MB  4.2MB/s   02:01    system01.dbf                                                                             100%  710MB   5.3MB/s  02:13    temp01.dbf                                                                                100%   29MB  5.8MB/s   00:05    undotbs01.dbf                                                                            100%   75MB   5.0MB/s  00:15    users01.dbf                                                                               100%5128KB   5.0MB/s   00:01 6.     将source 端的初始化文件和口令文件copy到target 端SQL> create pfile from spfile;File created. [oracle@tianlesoftware dbs]$ scp initanqing.ora 192.168.3.201:/u02/app/oracle/product/11.2.0/db_1/dbsoracle@192.168.3.201"s password: initanqing.ora                                                                           100% 1017     1.0KB/s   00:00  7.     在Target 端修改初始化文件修改的内容包括控制文件保存位置，user_dump_dest,background_dump_dest,core_dump_dest 等。 修改之后如下：rac1:/u02/app/oracle/product/11.2.0/db_1/dbs>cat initanqing.ora anqing.__db_cache_size=570425344anqing.__java_pool_size=16777216anqing.__large_pool_size=16777216anqing.__oracle_base="/u02/app/oracle"#ORACLE_BASEset from environmentanqing.__pga_aggregate_target=570425344anqing.__sga_target=838860800anqing.__shared_io_pool_size=0anqing.__shared_pool_size=218103808anqing.__streams_pool_size=0*.audit_file_dest="/u02/app/oracle/admin/anqing/adump"*.audit_trail="db"*.compatible="11.2.0.0.0"*.control_files="/u02/app/oracle/oradata/anqing/control01.ctl","/u02/app/oracle/fast_recovery_area/anqing/control02.ctl"*.db_block_size=8192*.db_domain=""*.db_name="anqing"*.db_recovery_file_dest="/u02/app/oracle/fast_recovery_area"*.db_recovery_file_dest_size=4194304000*.diagnostic_dest="/u02/app/oracle"*.dispatchers="（PROTOCOL=TCP）（SERVICE=anqingXDB）"*.log_archive_dest_1="LOCATION=/u02/archivelog"*.log_archive_format="%t_%s_%r.arc"*.memory_target=1393557504*.open_cursors=300*.processes=150*.remote_login_passwordfile="EXCLUSIVE"*.undo_tablespace="UNDOTBS1" 8.     在Target 端重建控制文件 在重建控制之前先rm 掉我们从Source 端copy 过来的控制文件，同时修改我们的控制文件代码，使路径使用Target 的路径。
rac1:/u02/app/oracle/oradata/anqing> lscontrol01.ctl  redo02.log sysaux01.dbf  temp01.dbf     users01.dbfredo01.log     redo03.log system01.dbf  undotbs01.dbfrac1:/u02/app/oracle/oradata/anqing> rmcontrol01.ctl rac1:/u02/app/oracle/oradata/anqing> rmtemp01.dbfrac1:/u02/app/oracle/oradata/anqing> lsredo01.log redo02.log  redo03.log  sysaux01.dbf system01.dbf  undotbs01.dbf  users01.dbf --使用pfile 启动DB 致nomout 状态：rac1:/u02/app/oracle/oradata/anqing>sqlplus / as sysdba SQL*Plus: Release 11.2.0.3.0 Production onMon Feb 13 21:20:08 2012Copyright （c） 1982, 2011, Oracle.  All rights reserved.Connected to an idle instance. SQL> startup nomount pfile="/u02/app/oracle/product/11.2.0/db_1/dbs/initanqing.ora";ORACLE instance started. Total System Global Area 1402982400 bytesFixed Size                  2228304 bytesVariable Size             822087600 bytesDatabase Buffers          570425344 bytesRedo Buffers                8241152 bytesSQL> --重建控制文件：SQL> CREATE CONTROLFILE REUSE DATABASE"ANQING" NORESETLOGS ARCHIVELOG   MAXLOGFILES 16 2    3      MAXLOGMEMBERS 3 4      MAXDATAFILES 100 5      MAXINSTANCES 8 6      MAXLOGHISTORY 292 7  LOGFILE 8    GROUP 1"/u02/app/oracle/oradata/anqing/redo01.log" SIZE 50M BLOCKSIZE 512, 9    GROUP 2"/u02/app/oracle/oradata/anqing/redo02.log" SIZE 50M BLOCKSIZE 512, 10   GROUP 3 "/u02/app/oracle/oradata/anqing/redo03.log"  SIZE 50M BLOCKSIZE 512 11  --STANDBY LOGFILE 12 DATAFILE 13   "/u02/app/oracle/oradata/anqing/system01.dbf", 14   "/u02/app/oracle/oradata/anqing/sysaux01.dbf", 15   "/u02/app/oracle/oradata/anqing/undotbs01.dbf", 16   "/u02/app/oracle/oradata/anqing/users01.dbf" 17 CHARACTER SET ZHS16GBK 18  ; Control file created. --打开DB:SQL> alter database open;Database altered. --尝试重建TEMP表空间，报错：SQL> ALTER TABLESPACE TEMP ADD TEMPFILE"/u02/app/oracle/oradata/anqing/temp01.dbf" SIZE 30408704  REUSE AUTOEXTEND ON NEXT 655360  MAXSIZE 32767M;ALTER TABLESPACE TEMP ADD TEMPFILE"/u02/app/oracle/oradata/anqing/temp01.dbf" SIZE 30408704  REUSE AUTOEXTEND ON NEXT 655360  MAXSIZE 32767M*ERROR at line 1:ORA-00604: error occurred at recursive SQLlevel 1ORA-06553: PLS-801: internal error [56327]--这里出错，是因为我们从32迁移到64位，还没有进行PL/SQL 模块的重建，等我们重建完成，在重建TEMP就没有问题了。 9.     在Target 端重新编译64位的PL/SQL模块 SQL> SPOOL mig32-64.log;
SQL> @$ORACLE_HOME/rdbms/admin/utlirp.sql ……--脚本大概执行2分钟DOC>###############################################################DOC>  utlirp.sql completed successfully. All PL/SQL objects in theDOC>  database have been invalidated.DOC>DOC>  Shut down and restart the database in normal mode and run utlrp.sql toDOC>  recompile invalid objects.DOC>###############################################################SQL>SQL> SPOOL OFF; --这里说的很清楚，所有的PL/SQL 对象都变成了无效，需要shut down 后，用normal 默认启动，并执行utlrp.sql脚本，验证无效对象个数：SQL> select count（*） from all_objectswhere status="INVALID"; COUNT（*）----------     9768 --编译无效对象：SQL> shutdown immediateDatabase closed.Database dismounted.ORACLE instance shut down. SQL> startupORACLE instance started.Total System Global Area 1402982400 bytesFixed Size                  2228304 bytesVariable Size             822087600 bytesDatabase Buffers          570425344 bytesRedo Buffers                8241152 bytesDatabase mounted.Database opened.SQL> @？/rdbms/admin/utlrp.sql;--脚本执行半个小时 SQL> select count（*） from all_objectswhere status="INVALID"; COUNT（*）----------     8054 --还是有很多无效对象，先放着，处理完Java在编译一次 10.  重建 javashared data objects （SRO）用SYS 用户执行如下脚本：
begin
  update obj$ set status=5 where obj#=（select obj# fromobj$,javasnm$ 
    where owner#=0 and type#=29 and short（+）=name and 
   nvl（longdbcs,name）="oracle/aurora/rdbms/Compiler"）;
  commit;
  declare
    cursor C1 is select
       "DROP JAVA DATA "" || u.name||""."" || o.name || """
       from obj$ o,user$ u where o.type#=56 andu.user#=o.owner#;
      ddl_statement varchar2（200）;
    iterations number;
    previous_iterations number;
    loop_count number;
    my_err     number;
  begin
    previous_iterations := 10000000;
    loop
      -- To make sure we eventually stop,pick a maxnumber of iterations
      select count（*） into iterations from obj$ wheretype#=56;
      exit when iterations=0 or iterations >=previous_iterations;
      previous_iterations := iterations;
      loop_count := 0;
      open C1;
      loop
        begin
          fetch C1 intoddl_statement;
          exit when C1%NOTFOUND orloop_count > iterations;
        exception when others then
           my_err := sqlcode;
           if my_err = -1555then --snapshot too old, re-execute fetch query
             exit;
           else
             raise;
           end if;
        end;
        initjvmaux.exec（ddl_statement）;
        loop_count := loop_count + 1;
      end loop;
      close C1;
    end loop;
  end;
  commit;
  initjvmaux.drp（"delete fromjava$policy$shared$table"）;
  update obj$ set status=1 where obj#=（select obj# fromobj$,javasnm$ 
    where owner#=0 and type#=29 and short（+）=name and 
   nvl（longdbcs,name）="oracle/aurora/rdbms/Compiler"）;
  commit;
end;
/
--2分钟左右
create or replace java system
/--十分钟左右  11.   在编译一次无效对象 SQL> @？/rdbms/admin/utlrp.sql;--脚本执行时会产生大量的online redo，所以要保证有足够的online redo 来切换。 --这里突然想到把第九步的编译无效对象省略，放到这里一起执行，应该可以节省半个小时，不过这里不能测试了。 在执行脚本的过程中，session异常中断，在alert log和 trace里都有：ORA-7445的错误：ORA-07445: exception encountered: core dump[__intel_new_memcpy（）+2164] [SIGSEGV] [ADDR:0x7F2F0CBBD2BF] [PC:0x47ED7D4][Address not mapped to object] []
----- Current SQL Statement for thissession （sql_id=2y0pxmcj6k00t） -----ALTER VIEW"OLAPSYS"."ALL$OLAP2_AW_CATALOGS" COMPILE        导致这个问题，是因为我们启动了OLAP，在32位转到64位时就会遇到这种错误。 12.  处理OLAP 问题--查看组件SQL> select comp_id,comp_name,version,status from dba_registry; COMP_ID     COMP_NAME                          VERSION      STATUS----------------------------------------------- ------------ ------------------OWB         OWB                                11.2.0.3.0   VALIDAPEX        Oracle Application Express         3.2.1.00.12  VALIDEM          Oracle Enterprise Manager           11.2.0.3.0   VALIDAMD         OLAP Catalog                       11.2.0.3.0   INVALIDSDO         Spatial                            11.2.0.3.0   INVALIDORDIM       Oracle Multimedia                  11.2.0.3.0   INVALIDXDB         Oracle XML Database                11.2.0.3.0   INVALIDCONTEXT     Oracle Text                        11.2.0.3.0   VALIDEXF         Oracle Expression Filter           11.2.0.3.0   INVALIDRUL         Oracle Rules Manager               11.2.0.3.0   INVALIDOWM         Oracle Workspace Manager           11.2.0.3.0   INVALID COMP_ID     COMP_NAME                          VERSION      STATUS----------------------------------------------- ------------ ------------------CATALOG     Oracle Database Catalog Views      11.2.0.3.0   INVALIDCATPROC     Oracle Database Packages and Types 11.2.0.3.0   INVALIDJAVAVM      JServer JAVA Virtual Machine       11.2.0.3.0   VALIDXML         Oracle XDK                         11.2.0.3.0   VALIDCATJAVA     Oracle Database Java Packages      11.2.0.3.0   VALIDAPS          OLAP Analytic Workspace             11.2.0.3.0   INVALIDXOQ         Oracle OLAP API                    11.2.0.3.0   VALID 18 rows selected. MOS 有篇文档提供了重建OLAP组件的方法：How To Remove Or To Reinstall The OLAPOption To 10g And 11g [ID 332351.1] 这里提供了2种解决方法：（1）在ORACLE_HOME 级别重建--删除OLAP备份DB和oraInventory,在关闭ORACLE_HOME上的所有Instance，使用Universal Installer （OUI） 移除ORACLE_HOME的OLAP 选项。在OUI 中选择升级已经存在的db，在selectoptions 选项可以控制是否启用OLAP。如下图： 
在Unix 和Linux 平台上，也可以直接relink olap_off 的方式来删除，这样就避免使用OUI。  cd $ORACLE_HOME/rdbms/lib make -f ins_rdbms.mk olap_off
 make -f ins_rdbms.mk ioracle        非注册的OLAP将不在db 的banner 中限制，同时v$option  也会显示为FALSE. --重新安装OLAP1）在OUI中添加OLAP 组件，参考上图。2）添加完毕之后执行如下脚本，在db 级别进行添加。sqlplus /nolog
SQL> conn /as sysdba
SQL> spool add_olap.log
SQL> @？/olap/admin/olap.sql SYSAUX TEMP;
SQL> @？/rdbms/admin/utlrp.sql
SQL> spool off （2）在Database 级别重建--删除OLAP --在db server 上开一个session，设定相关的变量，并用sys 用户执行如下脚本：ORACLE_HOME=your_oracle_home
ORACLE_SID=your_db_sid
PATH=$ORACLE_HOME/bin:$PATH
sqlplus /nolog
SQL> conn / as sysdba
----> Remove OLAP Catalog
SQL> @？/olap/admin/catnoamd.sql
----> Remove OLAP API
SQL> @？/olap/admin/olapidrp.plb
SQL> @？/olap/admin/catnoxoq.sql
----> Deinstall APS - OLAP AW component
SQL> @？/olap/admin/catnoaps.sql
SQL> @？/olap/admin/cwm2drop.sql
----> Recompile invalids
SQL> @？/rdbms/admin/utlrp.sql catnoamd.sqlused above will drop the OLAPSYS schema （which is completely OLAP specific）.--catnoamd.sql 脚本用来drop OLAPSYS 用户。cwm2drop.sqlneeds to be run only in 11g. In 10g, catnoamd.sql already calls it.--cwm2drop.sql 只需要在11g中执行。在10g中，catnoamd.sql 脚本中已经包含了该脚本，所以不用执行。 If you connectto the database in SQLPLUS, the banner still shows the OLAP option,this happensbecause the Oracle executable is still aware of OLAP, it is linked with OLAP, andalso the OLAP files are still present in the ORACLE_HOME. Once running thedeinstall scripts, the utlrp.sql could report invalid objects with SYS andPUBLIC owners. If so, the following SQL can be used to report more details onthat:--删除之后，可能导致一些无效的对象，可以使用如下SQL 来查看具体信息： SQL> select owner, object_name,object_type, status from dba_objects where status="INVALID"; You could getsome invalid objects under SYS and PUBLIC owner, and they are the old duplicate OLAPSYS objects copied under these schemas when Olap has been installedpreviously.Generally, theOlap objects are named with context like %OLAP%, %AWM%, or other Olap word keys intheir "object_name" field, however, if it needs a help to recognizethem, then please contact Oracle  Support and create a Service Request toget assistance for this question. Note that catnoadm.sql could fail from 10.1.0.5 to 11.1.0.7release.  Due to the factthat it refers to three scripts which don"t get shipped until 11.2 this scriptwill fail.Besides that itwill error on 7 non-existing synonyms to drop.
Prior 11.2, execute these three dropsynonym statements:
SQL> drop public synonym OlapFactView;
SQL> drop public synonym OlapDimView;
SQL> drop public synonym DBMS_ODM; If there arefurther invalid OLAP specific objects after the above removal steps, pleaseconsult Note1060023.1, Note565773.1 

If you do notwant to install OLAP into a newly created database, then in the DatabaseConfiguration Assistant （dbca）, select a custom database template and uncheckOracle OLAP in the Database Components wizard step.
--添加OLAPAssuming thatyou created your database manually or via DBCA, add the OLAP option to anexisting Enterprise Edition Database.--假设我们在ORACLE_HOME 级别添加了OLAP。Then open aterminal session on the database server machine, set ORACLE_HOME, ORACLE_SIDenvironment variables, and execute some script as user SYS AS SYSDBA, detailsbelow:--在db server 端用SYS 用户执行如下脚本，来创建olap。 ORACLE_HOME=your_oracle_home
ORACLE_SID=your_db_sid
PATH=$ORACLE_HOME/bin:$PATH
sqlplus /nolog
SQL> conn /as sysdba
SQL> @？/olap/admin/olap.sql SYSAUX TEMP;--这个脚本需要30+分钟 我这里按照DATABASE 的级别重建了OLAP。 13.  验证无效对象和组件状态在我们重建OLAP 之前有很多的无效组件，现在查看，都变成了有效状态，并且也没有了无效对象，如果到这一步还有无效对象，那么继续执行utlrp.sql 来编译。 SQL> select count（*） from all_objectswhere status="INVALID"; COUNT（*）----------        0 SQL> selectcomp_id,comp_name,version,status from dba_registry;COMP_ID     COMP_NAME                          VERSION      STATUS----------------------------------------------- ------------ ------------------AMD         OLAP Catalog                       11.2.0.3.0   VALIDOWB         OWB                                11.2.0.3.0   VALIDAPEX        Oracle Application Express         3.2.1.00.12  VALIDEM          Oracle Enterprise Manager          11.2.0.3.0   VALIDSDO         Spatial                            11.2.0.3.0   INVALIDORDIM       Oracle Multimedia                  11.2.0.3.0   VALIDXDB         Oracle XML Database                11.2.0.3.0   VALIDCONTEXT     Oracle Text                        11.2.0.3.0   VALIDEXF         Oracle Expression Filter           11.2.0.3.0   VALIDRUL         Oracle Rules Manager               11.2.0.3.0   VALIDOWM          Oracle Workspace Manager            11.2.0.3.0   VALID COMP_ID     COMP_NAME                          VERSION      STATUS----------------------------------------------- ------------ ------------------CATALOG     Oracle Database Catalog Views      11.2.0.3.0   VALIDCATPROC     Oracle Database Packages and Types 11.2.0.3.0   VALIDJAVAVM      JServer JAVA Virtual Machine       11.2.0.3.0   VALIDXML         Oracle XDK                         11.2.0.3.0   VALIDCATJAVA     Oracle Database Java Packages      11.2.0.3.0   VALIDAPS         OLAP Analytic Workspace            11.2.0.3.0   VALIDXOQ         Oracle OLAP API                    11.2.0.3.0   VALID18 rows selected.这里还有一个组件：SDO 是无效的，这个不在这篇的讨论范围。之前有整理过一篇组件重建的Blog，不过这篇条理很不清楚，回头重新整理一篇ORACLE 组件重建的blog。MDSYSSpatial De-install http://www.linuxidc.com/Linux/2012-03/55760.htm14.重建临时表空间，创建spfile 文件SQL> ALTER TABLESPACE TEMP ADD TEMPFILE"/u02/app/oracle/oradata/anqing/temp01.dbf" SIZE 30408704  REUSE AUTOEXTEND ON NEXT 655360  MAXSIZE 32767M;Tablespace altered.--编译过64位的PL/SQL对象以后，就可以正常创建了。SQL> create spfile from pfile;File created. SQL> select * from v$version; BANNER--------------------------------------------------------------------------------Oracle Database 11g Enterprise EditionRelease 11.2.0.3.0 - 64bit ProductionPL/SQL Release 11.2.0.3.0 - ProductionCORE   11.2.0.3.0      ProductionTNS for Linux: Version 11.2.0.3.0 -ProductionNLSRTL Version 11.2.0.3.0 - Production至此，迁移操作结束。
  • 1
  • 2
  • 3
  • 下一页
【内容导航】
第1页：直接 copy datafiles第2页：使用 RMAN Restore 示例
第3页：使用导出导入 说明
64位 Linux平台下Oracle安装文档Oracle 使用RMAN 将 DB 从10g 直接 Restore 到11g 示例相关资讯      oracle数据库教程 
  • Oracle raw数据类型介绍  （01/29/2013 10:05:53）
  • 监听器注册与ORA-12514 错误分析  （11/13/2012 14:30:08）
  • Oracle SQL的cursor理解  （11/13/2012 14:16:17）
  • Oracle 如何强制刷新Buffer Cache  （01/29/2013 10:02:46）
  • dblink致Oracle库的SCN变成两库的  （11/13/2012 14:24:41）
  • Linux操作系统下完全删除Oracle数  （11/13/2012 08:25:52）
本文评论 查看全部评论 （0）
表情： 姓名： 字数