简体中文版经机器翻译而成，仅供参考。如与英语版出现任何冲突，应以英语版为准。

通过日志传送实现Oracle数据库迁移

04/08/2024 贡献者

使用日志传送进行迁移的目标是、在新位置创建原始数据文件的副本、然后建立将更改传送到新环境的方法。

建立日志后、可以自动进行日志传输和重放、以使副本数据库与源数据库大致保持同步。例如、可以计划cron作业：(a)将最新日志复制到新位置、(b)每15分钟重放一次。这样做可以最大程度地减少转换时的中断、因为回写的归档日志不能超过15分钟。

下面显示的操作步骤本质上也是一个数据库克隆操作。显示的逻辑类似于NetApp SnapManager for Oracle (SMO)和NetApp SnapCenter Oracle插件中的引擎。某些客户已使用脚本或WFA工作流中显示的操作步骤执行自定义克隆操作。虽然此操作步骤比使用SMO或SnapCenter更需要手动操作、但仍可随时编写脚本、ONTAP中的数据管理API进一步简化了此过程。

日志传送-文件系统到文件系统

此示例演示了将名为waffle的数据库从普通文件系统迁移到位于不同服务器上的另一个普通文件系统的过程。同时、还展示了如何使用SnapMirror快速复制数据文件、但这并不是整个操作步骤不可或缺的一部分。

创建数据库备份

第一步是创建数据库备份。具体来说、此操作步骤需要一组数据文件、可用于归档日志重放。

environment

在此示例中、源数据库位于ONTAP系统上。创建数据库备份的最简单方法是使用快照。数据库将处于热备份模式几秒钟、而处于 snapshot create 在托管数据文件的卷上执行此操作。

SQL> alter database begin backup;
Database altered.

Cluster01::*> snapshot create -vserver vserver1 -volume jfsc1_oradata hotbackup
Cluster01::*>

SQL> alter database end backup;
Database altered.

结果是在磁盘上生成一个名为的快照 hotbackup 该映像包含处于热备份模式时的数据文件映像。如果将此快照中的数据与相应的归档日志结合使用以使数据文件保持一致、则可以将此快照中的数据用作还原或克隆的基础。在这种情况下、它会复制到新服务器。

还原到新环境

现在、必须在新环境中还原备份。这可以通过多种方式实现、包括Oracle RMAN、从备份应用程序(如NetBackup)还原、或者对处于热备份模式的数据文件执行简单的复制操作。

在此示例中、使用SnapMirror将快照热备份复制到新位置。

创建新卷以接收快照数据。从初始化镜像 jfsc1_oradata to vol_oradata。

Cluster01::*> volume create -vserver vserver1 -volume vol_oradata -aggregate data_01 -size 20g -state online -type DP -snapshot-policy none -policy jfsc3
[Job 833] Job succeeded: Successful

Cluster01::*> snapmirror initialize -source-path vserver1:jfsc1_oradata -destination-path vserver1:vol_oradata
Operation is queued: snapmirror initialize of destination "vserver1:vol_oradata".
Cluster01::*> volume mount -vserver vserver1 -volume vol_oradata -junction-path /vol_oradata
Cluster01::*>

在SnapMirror设置状态(指示同步已完成)后、请根据所需的快照专门更新镜像。

Cluster01::*> snapmirror show -destination-path vserver1:vol_oradata -fields state
source-path             destination-path        state
----------------------- ----------------------- ------------
vserver1:jfsc1_oradata vserver1:vol_oradata SnapMirrored

Cluster01::*> snapmirror update -destination-path vserver1:vol_oradata -source-snapshot hotbackup
Operation is queued: snapmirror update of destination "vserver1:vol_oradata".

可以通过查看来验证同步是否成功 newest-snapshot 字段。

Cluster01::*> snapmirror show -destination-path vserver1:vol_oradata -fields newest-snapshot
source-path             destination-path        newest-snapshot
----------------------- ----------------------- ---------------
vserver1:jfsc1_oradata vserver1:vol_oradata hotbackup

然后、可以断开镜像。

Cluster01::> snapmirror break -destination-path vserver1:vol_oradata
Operation succeeded: snapmirror break for destination "vserver1:vol_oradata".
Cluster01::>

挂载新文件系统。对于基于块的文件系统、具体过程因使用的LVM而异。必须配置FC分区或iSCSI连接。与LUN建立连接后、可以使用Linux等命令 pvscan 可能需要查找哪些卷组或LUN需要正确配置才能被ASM发现。

在此示例中、使用的是简单的NFS文件系统。可以直接挂载此文件系统。
```
fas8060-nfs1:/vol_oradata        19922944   1639360   18283584   9% /oradata
fas8060-nfs1:/vol_logs            9961472       128    9961344   1% /logs
```

创建控制文件创建模板

接下来必须创建控制文件模板。。 backup controlfile to trace 命令用于创建文本命令以重新创建控制文件。在某些情况下、此功能对于从备份还原数据库非常有用、并且通常与执行数据库克隆等任务的脚本结合使用。

以下命令的输出用于为迁移的数据库重新创建控制文件。

SQL> alter database backup controlfile to trace as '/tmp/waffle.ctrl';
Database altered.

创建控制文件后，将文件复制到新服务器。

[oracle@jfsc3 tmp]$ scp oracle@jfsc1:/tmp/waffle.ctrl /tmp/
oracle@jfsc1's password:
waffle.ctrl                                              100% 5199     5.1KB/s   00:00

备份参数文件

在新环境中、还需要一个参数文件。最简单的方法是从当前的spfile或pfile创建一个pfile。在此示例中、源数据库使用的是spfile。

SQL> create pfile='/tmp/waffle.tmp.pfile' from spfile;
File created.

创建oratab条目

要使oraenv等实用程序正常运行、必须创建oratab条目。要创建oratab条目、请完成以下步骤。

WAFFLE:/orabin/product/12.1.0/dbhome_1:N

准备目录结构

如果所需目录不存在、则必须创建它们、否则数据库启动操作步骤将失败。要准备目录结构、请满足以下最低要求。

[oracle@jfsc3 ~]$ . oraenv
ORACLE_SID = [oracle] ? WAFFLE
The Oracle base has been set to /orabin
[oracle@jfsc3 ~]$ cd $ORACLE_BASE
[oracle@jfsc3 orabin]$ cd admin
[oracle@jfsc3 admin]$ mkdir WAFFLE
[oracle@jfsc3 admin]$ cd WAFFLE
[oracle@jfsc3 WAFFLE]$ mkdir adump dpdump pfile scripts xdb_wallet

参数文件更新

要将参数文件复制到新服务器、请运行以下命令。默认位置为 $ORACLE_HOME/dbs 目录。在这种情况下、pfile可以放置在任何位置。它仅用作迁移过程中的中间步骤。

[oracle@jfsc3 admin]$ scp oracle@jfsc1:/tmp/waffle.tmp.pfile $ORACLE_HOME/dbs/waffle.tmp.pfile
oracle@jfsc1's password:
waffle.pfile                                             100%  916     0.9KB/s   00:00

根据需要编辑文件。例如、如果归档日志位置已更改、则必须更改pfile以反映新位置。在此示例中、仅重新定位控制文件、部分目的是在日志和数据文件系统之间分布控制文件。

[root@jfsc1 tmp]# cat waffle.pfile
WAFFLE.__data_transfer_cache_size=0
WAFFLE.__db_cache_size=507510784
WAFFLE.__java_pool_size=4194304
WAFFLE.__large_pool_size=20971520
WAFFLE.__oracle_base='/orabin'#ORACLE_BASE set from environment
WAFFLE.__pga_aggregate_target=268435456
WAFFLE.__sga_target=805306368
WAFFLE.__shared_io_pool_size=29360128
WAFFLE.__shared_pool_size=234881024
WAFFLE.__streams_pool_size=0
*.audit_file_dest='/orabin/admin/WAFFLE/adump'
*.audit_trail='db'
*.compatible='12.1.0.2.0'
*.control_files='/oradata//WAFFLE/control01.ctl','/oradata//WAFFLE/control02.ctl'
*.control_files='/oradata/WAFFLE/control01.ctl','/logs/WAFFLE/control02.ctl'
*.db_block_size=8192
*.db_domain=''
*.db_name='WAFFLE'
*.diagnostic_dest='/orabin'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=WAFFLEXDB)'
*.log_archive_dest_1='LOCATION=/logs/WAFFLE/arch'
*.log_archive_format='%t_%s_%r.dbf'
*.open_cursors=300
*.pga_aggregate_target=256m
*.processes=300
*.remote_login_passwordfile='EXCLUSIVE'
*.sga_target=768m
*.undo_tablespace='UNDOTBS1'

编辑完成后、根据此pfile创建一个spfile。

SQL> create spfile from pfile='waffle.tmp.pfile';
File created.

重新创建控制文件

在上一步中、是的输出 backup controlfile to trace 已复制到新服务器。所需输出的具体部分是 controlfile recreation 命令：此信息可在标记的部分下的文件中找到 Set #1. NORESETLOGS。它从行开始 create controlfile reuse database 并应包含该词 noresetlogs。以分号(;)字符结尾。

在此示例操作步骤中、该文件如下所示。

CREATE CONTROLFILE REUSE DATABASE "WAFFLE" NORESETLOGS  ARCHIVELOG
    MAXLOGFILES 16
    MAXLOGMEMBERS 3
    MAXDATAFILES 100
    MAXINSTANCES 8
    MAXLOGHISTORY 292
LOGFILE
  GROUP 1 '/logs/WAFFLE/redo/redo01.log'  SIZE 50M BLOCKSIZE 512,
  GROUP 2 '/logs/WAFFLE/redo/redo02.log'  SIZE 50M BLOCKSIZE 512,
  GROUP 3 '/logs/WAFFLE/redo/redo03.log'  SIZE 50M BLOCKSIZE 512
-- STANDBY LOGFILE
DATAFILE
  '/oradata/WAFFLE/system01.dbf',
  '/oradata/WAFFLE/sysaux01.dbf',
  '/oradata/WAFFLE/undotbs01.dbf',
  '/oradata/WAFFLE/users01.dbf'
CHARACTER SET WE8MSWIN1252
;

根据需要编辑此脚本、以反映各种文件的新位置。例如、某些已知支持高I/O的数据文件可能会重定向到高性能存储层上的文件系统。在其他情况下、更改可能纯粹出于管理员原因、例如、将给定PDB的数据文件隔离到专用卷中。

在此示例中、将显示 DATAFILE 虽然保持不变、但重做日志会移动到中的新位置 /redo 而不是与归档登录共享空间 /logs。

CREATE CONTROLFILE REUSE DATABASE "WAFFLE" NORESETLOGS  ARCHIVELOG
    MAXLOGFILES 16
    MAXLOGMEMBERS 3
    MAXDATAFILES 100
    MAXINSTANCES 8
    MAXLOGHISTORY 292
LOGFILE
  GROUP 1 '/redo/redo01.log'  SIZE 50M BLOCKSIZE 512,
  GROUP 2 '/redo/redo02.log'  SIZE 50M BLOCKSIZE 512,
  GROUP 3 '/redo/redo03.log'  SIZE 50M BLOCKSIZE 512
-- STANDBY LOGFILE
DATAFILE
  '/oradata/WAFFLE/system01.dbf',
  '/oradata/WAFFLE/sysaux01.dbf',
  '/oradata/WAFFLE/undotbs01.dbf',
  '/oradata/WAFFLE/users01.dbf'
CHARACTER SET WE8MSWIN1252
;

SQL> startup nomount;
ORACLE instance started.
Total System Global Area  805306368 bytes
Fixed Size                  2929552 bytes
Variable Size             331353200 bytes
Database Buffers          465567744 bytes
Redo Buffers                5455872 bytes
SQL> CREATE CONTROLFILE REUSE DATABASE "WAFFLE" NORESETLOGS  ARCHIVELOG
  2      MAXLOGFILES 16
  3      MAXLOGMEMBERS 3
  4      MAXDATAFILES 100
  5      MAXINSTANCES 8
  6      MAXLOGHISTORY 292
  7  LOGFILE
  8    GROUP 1 '/redo/redo01.log'  SIZE 50M BLOCKSIZE 512,
  9    GROUP 2 '/redo/redo02.log'  SIZE 50M BLOCKSIZE 512,
 10    GROUP 3 '/redo/redo03.log'  SIZE 50M BLOCKSIZE 512
 11  -- STANDBY LOGFILE
 12  DATAFILE
 13    '/oradata/WAFFLE/system01.dbf',
 14    '/oradata/WAFFLE/sysaux01.dbf',
 15    '/oradata/WAFFLE/undotbs01.dbf',
 16    '/oradata/WAFFLE/users01.dbf'
 17  CHARACTER SET WE8MSWIN1252
 18  ;
Control file created.
SQL>

如果任何文件放错位置或参数配置错误、则会生成错误、指示必须修复的问题。数据库已挂载、但尚未打开、无法打开、因为正在使用的数据文件仍标记为处于热备份模式。必须先应用归档日志、以使数据库保持一致。

初始日志复制

要使数据文件保持一致、至少需要执行一个日志回复操作。有许多选项可用于重放日志。在某些情况下、可以通过NFS共享原始服务器上的原始归档日志位置、并且可以直接进行日志回复。在其他情况下、必须复制归档日志。

例如、一个简单的 scp 此操作可以将所有当前日志从源服务器复制到迁移服务器：

[oracle@jfsc3 arch]$ scp jfsc1:/logs/WAFFLE/arch/* ./
oracle@jfsc1's password:
1_22_912662036.dbf                                       100%   47MB  47.0MB/s   00:01
1_23_912662036.dbf                                       100%   40MB  40.4MB/s   00:00
1_24_912662036.dbf                                       100%   45MB  45.4MB/s   00:00
1_25_912662036.dbf                                       100%   41MB  40.9MB/s   00:01
1_26_912662036.dbf                                       100%   39MB  39.4MB/s   00:00
1_27_912662036.dbf                                       100%   39MB  38.7MB/s   00:00
1_28_912662036.dbf                                       100%   40MB  40.1MB/s   00:01
1_29_912662036.dbf                                       100%   17MB  16.9MB/s   00:00
1_30_912662036.dbf                                       100%  636KB 636.0KB/s   00:00

初始日志重放

文件位于归档日志位置后、可以发出命令来重新显示它们 recover database until cancel 然后是响应 AUTO 自动重放所有可用日志。

SQL> recover database until cancel;
ORA-00279: change 382713 generated at 05/24/2016 09:00:54 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_23_912662036.dbf
ORA-00280: change 382713 for thread 1 is in sequence #23
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
AUTO
ORA-00279: change 405712 generated at 05/24/2016 15:01:05 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_24_912662036.dbf
ORA-00280: change 405712 for thread 1 is in sequence #24
ORA-00278: log file '/logs/WAFFLE/arch/1_23_912662036.dbf' no longer needed for
this recovery
...
ORA-00279: change 713874 generated at 05/26/2016 04:26:43 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_31_912662036.dbf
ORA-00280: change 713874 for thread 1 is in sequence #31
ORA-00278: log file '/logs/WAFFLE/arch/1_30_912662036.dbf' no longer needed for
this recovery
ORA-00308: cannot open archived log '/logs/WAFFLE/arch/1_31_912662036.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3

最终归档日志回复报告错误、但这是正常的。日志指示 sqlplus 正在查找特定日志文件、但未找到它。原因很可能是日志文件尚不存在。

如果可以在复制归档日志之前关闭源数据库、则只能执行此步骤一次。归档日志会进行复制和重做、然后、该过程可以直接继续执行转换过程、以复制关键重做日志。

增量日志复制和重放

在大多数情况下、不会立即执行迁移。迁移过程可能需要几天甚至几周才能完成、这意味着必须将日志持续运送到副本数据库并进行重新显示。因此、在转换完成后、必须传输和回显示最少的数据。

这样做可以通过多种方式编写脚本、但更常见的方法之一是使用rsync、这是一个常见的文件复制实用程序。使用此实用程序的最安全方法是将其配置为守护进程。例如、 rsyncd.conf 下面的文件显示了如何创建名为的资源 waffle.arch 可通过Oracle用户凭据访问并映射到 /logs/WAFFLE/arch。最重要的是、资源设置为只读、这样可以读取生产数据、但不会对其进行更改。

[root@jfsc1 arch]# cat /etc/rsyncd.conf
[waffle.arch]
   uid=oracle
   gid=dba
   path=/logs/WAFFLE/arch
   read only = true
[root@jfsc1 arch]# rsync --daemon

以下命令将新服务器的归档日志目标与rsync资源同步 waffle.arch 在原始服务器上。。 t 中的参数 rsync - potg 根据时间戳比较文件列表、并且仅复制新文件。此过程会对新服务器进行增量更新。也可以在cron中计划定期运行此命令。

[oracle@jfsc3 arch]$ rsync -potg --stats --progress jfsc1::waffle.arch/* /logs/WAFFLE/arch/
1_31_912662036.dbf
      650240 100%  124.02MB/s    0:00:00 (xfer#1, to-check=8/18)
1_32_912662036.dbf
     4873728 100%  110.67MB/s    0:00:00 (xfer#2, to-check=7/18)
1_33_912662036.dbf
     4088832 100%   50.64MB/s    0:00:00 (xfer#3, to-check=6/18)
1_34_912662036.dbf
     8196096 100%   54.66MB/s    0:00:00 (xfer#4, to-check=5/18)
1_35_912662036.dbf
    19376128 100%   57.75MB/s    0:00:00 (xfer#5, to-check=4/18)
1_36_912662036.dbf
       71680 100%  201.15kB/s    0:00:00 (xfer#6, to-check=3/18)
1_37_912662036.dbf
     1144320 100%    3.06MB/s    0:00:00 (xfer#7, to-check=2/18)
1_38_912662036.dbf
    35757568 100%   63.74MB/s    0:00:00 (xfer#8, to-check=1/18)
1_39_912662036.dbf
      984576 100%    1.63MB/s    0:00:00 (xfer#9, to-check=0/18)
Number of files: 18
Number of files transferred: 9
Total file size: 399653376 bytes
Total transferred file size: 75143168 bytes
Literal data: 75143168 bytes
Matched data: 0 bytes
File list size: 474
File list generation time: 0.001 seconds
File list transfer time: 0.000 seconds
Total bytes sent: 204
Total bytes received: 75153219
sent 204 bytes  received 75153219 bytes  150306846.00 bytes/sec
total size is 399653376  speedup is 5.32

收到日志后、必须对其进行重新显示。前面的示例显示了如何使用sqlplus手动运行 recover database until cancel，一个可以轻松实现自动化的过程。此处显示的示例使用中所述的脚本 "重放数据库上的日志"。这些脚本接受一个参数、用于指定需要重放操作的数据库。这样就可以在多数据库迁移工作中使用相同的脚本。

[oracle@jfsc3 logs]$ ./replay.logs.pl WAFFLE
ORACLE_SID = [WAFFLE] ? The Oracle base remains unchanged with value /orabin
SQL*Plus: Release 12.1.0.2.0 Production on Thu May 26 10:47:16 2016
Copyright (c) 1982, 2014, Oracle.  All rights reserved.
Connected to:
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options
SQL> ORA-00279: change 713874 generated at 05/26/2016 04:26:43 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_31_912662036.dbf
ORA-00280: change 713874 for thread 1 is in sequence #31
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
ORA-00279: change 814256 generated at 05/26/2016 04:52:30 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_32_912662036.dbf
ORA-00280: change 814256 for thread 1 is in sequence #32
ORA-00278: log file '/logs/WAFFLE/arch/1_31_912662036.dbf' no longer needed for
this recovery
ORA-00279: change 814780 generated at 05/26/2016 04:53:04 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_33_912662036.dbf
ORA-00280: change 814780 for thread 1 is in sequence #33
ORA-00278: log file '/logs/WAFFLE/arch/1_32_912662036.dbf' no longer needed for
this recovery
...
ORA-00279: change 1120099 generated at 05/26/2016 09:59:21 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_40_912662036.dbf
ORA-00280: change 1120099 for thread 1 is in sequence #40
ORA-00278: log file '/logs/WAFFLE/arch/1_39_912662036.dbf' no longer needed for
this recovery
ORA-00308: cannot open archived log '/logs/WAFFLE/arch/1_40_912662036.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
SQL> Disconnected from Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options

转换

准备好转换到新环境后、必须执行一次最终同步、其中包括归档日志和重做日志。如果原始重做日志位置尚不已知、则可按如下方式进行标识：

SQL> select member from v$logfile;
MEMBER
--------------------------------------------------------------------------------
/logs/WAFFLE/redo/redo01.log
/logs/WAFFLE/redo/redo02.log
/logs/WAFFLE/redo/redo03.log

关闭源数据库。
使用所需的方法在新服务器上对归档日志执行一次最终同步。

必须将源重做日志复制到新服务器。在此示例中、重做日志已重新定位到的新目录中 /redo。

[oracle@jfsc3 logs]$ scp jfsc1:/logs/WAFFLE/redo/* /redo/
oracle@jfsc1's password:
redo01.log                                                              100%   50MB  50.0MB/s   00:01
redo02.log                                                              100%   50MB  50.0MB/s   00:00
redo03.log                                                              100%   50MB  50.0MB/s   00:00

在此阶段、新数据库环境包含将其恢复到与源完全相同状态所需的所有文件。归档日志必须最后一次重新显示。

SQL> recover database until cancel;
ORA-00279: change 1120099 generated at 05/26/2016 09:59:21 needed for thread 1
ORA-00289: suggestion : /logs/WAFFLE/arch/1_40_912662036.dbf
ORA-00280: change 1120099 for thread 1 is in sequence #40
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
AUTO
ORA-00308: cannot open archived log '/logs/WAFFLE/arch/1_40_912662036.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
ORA-00308: cannot open archived log '/logs/WAFFLE/arch/1_40_912662036.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3

完成后、必须重做日志。如果消息 Media recovery complete 将返回、此过程将成功、数据库将同步并可打开。
```
SQL> recover database;
Media recovery complete.
SQL> alter database open;
Database altered.
```

日志传送- ASM到文件系统

此示例演示了如何使用Oracle RMAN迁移数据库。它与前面的文件系统到文件系统日志传送示例非常相似、但主机无法识别ASM上的文件。迁移ASM设备上的数据的唯一方法是重新定位ASM LUN或使用Oracle RMAN执行复制操作。

虽然从Oracle ASM复制文件时需要使用RMAN、但RMAN的使用并不限于ASM。RMAN可用于从任何类型的存储迁移到任何其他类型。

此示例显示了将名为pancake的数据库从ASM存储重新定位到位于路径不同服务器上的常规文件系统 /oradata 和 /logs。

创建数据库备份

第一步是为要迁移到备用服务器的数据库创建备份。由于源使用Oracle ASM、因此必须使用RMAN。可以按如下所示执行简单的RMAN备份。此方法会创建一个带标记的备份、稍后可通过RMAN在操作步骤中轻松识别该备份。

第一个命令用于定义备份的目标类型以及要使用的位置。第二个选项仅启动数据文件的备份。

RMAN> configure channel device type disk format '/rman/pancake/%U';
using target database control file instead of recovery catalog
old RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/%U';
new RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/%U';
new RMAN configuration parameters are successfully stored
RMAN> backup database tag 'ONTAP_MIGRATION';
Starting backup at 24-MAY-16
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=251 device type=DISK
channel ORA_DISK_1: starting full datafile backup set
channel ORA_DISK_1: specifying datafile(s) in backup set
input datafile file number=00001 name=+ASM0/PANCAKE/system01.dbf
input datafile file number=00002 name=+ASM0/PANCAKE/sysaux01.dbf
input datafile file number=00003 name=+ASM0/PANCAKE/undotbs101.dbf
input datafile file number=00004 name=+ASM0/PANCAKE/users01.dbf
channel ORA_DISK_1: starting piece 1 at 24-MAY-16
channel ORA_DISK_1: finished piece 1 at 24-MAY-16
piece handle=/rman/pancake/1gr6c161_1_1 tag=ONTAP_MIGRATION comment=NONE
channel ORA_DISK_1: backup set complete, elapsed time: 00:00:03
channel ORA_DISK_1: starting full datafile backup set
channel ORA_DISK_1: specifying datafile(s) in backup set
including current control file in backup set
including current SPFILE in backup set
channel ORA_DISK_1: starting piece 1 at 24-MAY-16
channel ORA_DISK_1: finished piece 1 at 24-MAY-16
piece handle=/rman/pancake/1hr6c164_1_1 tag=ONTAP_MIGRATION comment=NONE
channel ORA_DISK_1: backup set complete, elapsed time: 00:00:01
Finished backup at 24-MAY-16

备份控制文件

稍后需要在的操作步骤中为备份控制文件 duplicate database 操作。

RMAN> backup current controlfile format '/rman/pancake/ctrl.bkp';
Starting backup at 24-MAY-16
using channel ORA_DISK_1
channel ORA_DISK_1: starting full datafile backup set
channel ORA_DISK_1: specifying datafile(s) in backup set
including current control file in backup set
channel ORA_DISK_1: starting piece 1 at 24-MAY-16
channel ORA_DISK_1: finished piece 1 at 24-MAY-16
piece handle=/rman/pancake/ctrl.bkp tag=TAG20160524T032651 comment=NONE
channel ORA_DISK_1: backup set complete, elapsed time: 00:00:01
Finished backup at 24-MAY-16

备份参数文件

在新环境中、还需要一个参数文件。最简单的方法是从当前的spfile或pfile创建一个pfile。在此示例中、源数据库使用spfile。

RMAN> create pfile='/rman/pancake/pfile' from spfile;
Statement processed

ASM文件重命名脚本

移动数据库时，控制文件中当前定义的几个文件位置会发生变化。以下脚本将创建一个RMAN脚本、以便于执行此过程。此示例显示了一个数据文件数量非常少的数据库、但数据库通常包含数百甚至数千个数据文件。

此脚本可在中找到 "ASM到文件系统名称转换" 它做了两件事。

首先、它会创建一个参数来重新定义重做日志位置、该位置称为 log_file_name_convert。它本质上是一个交替字段的列表。第一个字段是当前重做日志的位置、第二个字段是新服务器上的位置。然后、重复执行此模式。

第二个功能是为数据文件重命名提供模板。该脚本循环显示数据文件、提取名称和文件编号信息、并将其格式化为RMAN脚本。然后、它会对临时文件执行相同的操作。结果是生成一个简单的RMAN脚本、可以根据需要进行编辑、以确保文件还原到所需位置。

SQL> @/rman/mk.rename.scripts.sql
Parameters for log file conversion:
*.log_file_name_convert = '+ASM0/PANCAKE/redo01.log',
'/NEW_PATH/redo01.log','+ASM0/PANCAKE/redo02.log',
'/NEW_PATH/redo02.log','+ASM0/PANCAKE/redo03.log', '/NEW_PATH/redo03.log'
rman duplication script:
run
{
set newname for datafile 1 to '+ASM0/PANCAKE/system01.dbf';
set newname for datafile 2 to '+ASM0/PANCAKE/sysaux01.dbf';
set newname for datafile 3 to '+ASM0/PANCAKE/undotbs101.dbf';
set newname for datafile 4 to '+ASM0/PANCAKE/users01.dbf';
set newname for tempfile 1 to '+ASM0/PANCAKE/temp01.dbf';
duplicate target database for standby backup location INSERT_PATH_HERE;
}
PL/SQL procedure successfully completed.

捕获此屏幕的输出。。 log_file_name_convert 参数将按如下所述放置在pfile中。必须相应地编辑RMAN数据文件重命名和重复脚本、才能将数据文件放置在所需位置。在此示例中、它们全部置于中 /oradata/pancake。

run
{
set newname for datafile 1 to '/oradata/pancake/pancake.dbf';
set newname for datafile 2 to '/oradata/pancake/sysaux.dbf';
set newname for datafile 3 to '/oradata/pancake/undotbs1.dbf';
set newname for datafile 4 to '/oradata/pancake/users.dbf';
set newname for tempfile 1 to '/oradata/pancake/temp.dbf';
duplicate target database for standby backup location '/rman/pancake';
}

准备目录结构

这些脚本几乎已准备就绪、可以执行、但首先必须设置好目录结构。如果所需目录不存在、则必须创建它们、否则数据库启动操作步骤将失败。以下示例反映了最低要求。

[oracle@jfsc2 ~]$ mkdir /oradata/pancake
[oracle@jfsc2 ~]$ mkdir /logs/pancake
[oracle@jfsc2 ~]$ cd /orabin/admin
[oracle@jfsc2 admin]$ mkdir PANCAKE
[oracle@jfsc2 admin]$ cd PANCAKE
[oracle@jfsc2 PANCAKE]$ mkdir adump dpdump pfile scripts xdb_wallet

创建oratab条目

要使oraenv等实用程序正常运行、需要使用以下命令。

PANCAKE:/orabin/product/12.1.0/dbhome_1:N

参数更新

必须更新保存的pfile、以反映新服务器上的任何路径更改。数据文件路径更改由RMAN复制脚本进行更改、几乎所有数据库都需要对进行更改 control_files 和 log_archive_dest parameters此外、还可能需要更改审核文件位置以及参数、例如 db_create_file_dest 在ASM之外可能不相关。经验丰富的DBA应在继续操作之前仔细查看建议的变更。

在此示例中、主要更改包括控制文件位置、日志归档目标以及的添加 log_file_name_convert 参数。

PANCAKE.__data_transfer_cache_size=0
PANCAKE.__db_cache_size=545259520
PANCAKE.__java_pool_size=4194304
PANCAKE.__large_pool_size=25165824
PANCAKE.__oracle_base='/orabin'#ORACLE_BASE set from environment
PANCAKE.__pga_aggregate_target=268435456
PANCAKE.__sga_target=805306368
PANCAKE.__shared_io_pool_size=29360128
PANCAKE.__shared_pool_size=192937984
PANCAKE.__streams_pool_size=0
*.audit_file_dest='/orabin/admin/PANCAKE/adump'
*.audit_trail='db'
*.compatible='12.1.0.2.0'
*.control_files='+ASM0/PANCAKE/control01.ctl','+ASM0/PANCAKE/control02.ctl'
*.control_files='/oradata/pancake/control01.ctl','/logs/pancake/control02.ctl'
*.db_block_size=8192
*.db_domain=''
*.db_name='PANCAKE'
*.diagnostic_dest='/orabin'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=PANCAKEXDB)'
*.log_archive_dest_1='LOCATION=+ASM1'
*.log_archive_dest_1='LOCATION=/logs/pancake'
*.log_archive_format='%t_%s_%r.dbf'
'/logs/path/redo02.log'
*.log_file_name_convert = '+ASM0/PANCAKE/redo01.log', '/logs/pancake/redo01.log', '+ASM0/PANCAKE/redo02.log', '/logs/pancake/redo02.log', '+ASM0/PANCAKE/redo03.log',  '/logs/pancake/redo03.log'
*.open_cursors=300
*.pga_aggregate_target=256m
*.processes=300
*.remote_login_passwordfile='EXCLUSIVE'
*.sga_target=768m
*.undo_tablespace='UNDOTBS1'

确认新参数后、必须将这些参数生效。虽然存在多个选项、但大多数客户都会根据文本pfile创建spfile。

bash-4.1$ sqlplus / as sysdba
SQL*Plus: Release 12.1.0.2.0 Production on Fri Jan 8 11:17:40 2016
Copyright (c) 1982, 2014, Oracle.  All rights reserved.
Connected to an idle instance.
SQL> create spfile from pfile='/rman/pancake/pfile';
File created.

启动非挂载

复制数据库前的最后一步是启动数据库进程、但不挂载文件。在此步骤中、spfile可能会出现明显问题。如果 startup nomount 命令因参数错误而失败、关闭、更正pfile模板、将其重新加载为spfile并重试非常简单。

SQL> startup nomount;
ORACLE instance started.
Total System Global Area  805306368 bytes
Fixed Size                  2929552 bytes
Variable Size             373296240 bytes
Database Buffers          423624704 bytes
Redo Buffers                5455872 bytes

复制数据库

与此过程中的其他步骤相比、将先前的RMAN备份还原到新位置所需的时间更长。必须在不更改数据库ID (DBID)或不重置日志的情况下复制数据库。这样可以防止应用日志、而这是完全同步副本所必需的步骤。

使用在上一步中创建的脚本、使用RMAN作为aux连接到数据库、并使用问题描述the DUKATE DATABASE命令。

[oracle@jfsc2 pancake]$ rman auxiliary /
Recovery Manager: Release 12.1.0.2.0 - Production on Tue May 24 03:04:56 2016
Copyright (c) 1982, 2014, Oracle and/or its affiliates.  All rights reserved.
connected to auxiliary database: PANCAKE (not mounted)
RMAN> run
2> {
3> set newname for datafile 1 to '/oradata/pancake/pancake.dbf';
4> set newname for datafile 2 to '/oradata/pancake/sysaux.dbf';
5> set newname for datafile 3 to '/oradata/pancake/undotbs1.dbf';
6> set newname for datafile 4 to '/oradata/pancake/users.dbf';
7> set newname for tempfile 1 to '/oradata/pancake/temp.dbf';
8> duplicate target database for standby backup location '/rman/pancake';
9> }
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
Starting Duplicate Db at 24-MAY-16
contents of Memory Script:
{
   restore clone standby controlfile from  '/rman/pancake/ctrl.bkp';
}
executing Memory Script
Starting restore at 24-MAY-16
allocated channel: ORA_AUX_DISK_1
channel ORA_AUX_DISK_1: SID=243 device type=DISK
channel ORA_AUX_DISK_1: restoring control file
channel ORA_AUX_DISK_1: restore complete, elapsed time: 00:00:01
output file name=/oradata/pancake/control01.ctl
output file name=/logs/pancake/control02.ctl
Finished restore at 24-MAY-16
contents of Memory Script:
{
   sql clone 'alter database mount standby database';
}
executing Memory Script
sql statement: alter database mount standby database
released channel: ORA_AUX_DISK_1
allocated channel: ORA_AUX_DISK_1
channel ORA_AUX_DISK_1: SID=243 device type=DISK
contents of Memory Script:
{
   set newname for tempfile  1 to
 "/oradata/pancake/temp.dbf";
   switch clone tempfile all;
   set newname for datafile  1 to
 "/oradata/pancake/pancake.dbf";
   set newname for datafile  2 to
 "/oradata/pancake/sysaux.dbf";
   set newname for datafile  3 to
 "/oradata/pancake/undotbs1.dbf";
   set newname for datafile  4 to
 "/oradata/pancake/users.dbf";
   restore
   clone database
   ;
}
executing Memory Script
executing command: SET NEWNAME
renamed tempfile 1 to /oradata/pancake/temp.dbf in control file
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
Starting restore at 24-MAY-16
using channel ORA_AUX_DISK_1
channel ORA_AUX_DISK_1: starting datafile backup set restore
channel ORA_AUX_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_AUX_DISK_1: restoring datafile 00001 to /oradata/pancake/pancake.dbf
channel ORA_AUX_DISK_1: restoring datafile 00002 to /oradata/pancake/sysaux.dbf
channel ORA_AUX_DISK_1: restoring datafile 00003 to /oradata/pancake/undotbs1.dbf
channel ORA_AUX_DISK_1: restoring datafile 00004 to /oradata/pancake/users.dbf
channel ORA_AUX_DISK_1: reading from backup piece /rman/pancake/1gr6c161_1_1
channel ORA_AUX_DISK_1: piece handle=/rman/pancake/1gr6c161_1_1 tag=ONTAP_MIGRATION
channel ORA_AUX_DISK_1: restored backup piece 1
channel ORA_AUX_DISK_1: restore complete, elapsed time: 00:00:07
Finished restore at 24-MAY-16
contents of Memory Script:
{
   switch clone datafile all;
}
executing Memory Script
datafile 1 switched to datafile copy
input datafile copy RECID=5 STAMP=912655725 file name=/oradata/pancake/pancake.dbf
datafile 2 switched to datafile copy
input datafile copy RECID=6 STAMP=912655725 file name=/oradata/pancake/sysaux.dbf
datafile 3 switched to datafile copy
input datafile copy RECID=7 STAMP=912655725 file name=/oradata/pancake/undotbs1.dbf
datafile 4 switched to datafile copy
input datafile copy RECID=8 STAMP=912655725 file name=/oradata/pancake/users.dbf
Finished Duplicate Db at 24-MAY-16

初始日志复制

现在、您必须将更改从源数据库发送到新位置。这样做可能需要多个步骤。最简单的方法是让源数据库上的RMAN将归档日志写出到共享网络连接。如果共享位置不可用、另一种方法是使用RMAN写入本地文件系统、然后使用rcp或rsync复制文件。

在此示例中、将显示 /rman 目录是一个NFS共享、可供原始数据库和迁移的数据库使用。

其中一个重要的问题描述是 disk format 条款。备份的磁盘格式为 %h_%e_%a.dbf，表示必须使用数据库的线程编号、序列号和激活ID格式。尽管字母不同、但这与匹配 log_archive_format='%t_%s_%r.dbf 参数。此参数还以线程编号、序列号和激活ID的格式指定归档日志。最终结果是、源上的日志文件备份会采用数据库预期的命名约定。这样做会执行等操作 recover database 更简单、因为sqlplus可以正确地预测要回显的归档日志的名称。

RMAN> configure channel device type disk format '/rman/pancake/logship/%h_%e_%a.dbf';
old RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/arch/%h_%e_%a.dbf';
new RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/logship/%h_%e_%a.dbf';
new RMAN configuration parameters are successfully stored
released channel: ORA_DISK_1
RMAN> backup as copy archivelog from time 'sysdate-2';
Starting backup at 24-MAY-16
current log archived
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=373 device type=DISK
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=54 RECID=70 STAMP=912658508
output file name=/rman/pancake/logship/1_54_912576125.dbf RECID=123 STAMP=912659482
channel ORA_DISK_1: archived log copy complete, elapsed time: 00:00:01
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=41 RECID=29 STAMP=912654101
output file name=/rman/pancake/logship/1_41_912576125.dbf RECID=124 STAMP=912659483
channel ORA_DISK_1: archived log copy complete, elapsed time: 00:00:01
...
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=45 RECID=33 STAMP=912654688
output file name=/rman/pancake/logship/1_45_912576125.dbf RECID=152 STAMP=912659514
channel ORA_DISK_1: archived log copy complete, elapsed time: 00:00:01
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=47 RECID=36 STAMP=912654809
output file name=/rman/pancake/logship/1_47_912576125.dbf RECID=153 STAMP=912659515
channel ORA_DISK_1: archived log copy complete, elapsed time: 00:00:01
Finished backup at 24-MAY-16

初始日志重放

文件位于归档日志位置后、可以发出命令来重新显示它们 recover database until cancel 然后是响应 AUTO 自动重放所有可用日志。参数文件当前正在将归档日志定向到 /logs/archive，但这与使用RMAN保存日志的位置不匹配。在恢复数据库之前、可以按如下所示临时重定向此位置。

SQL> alter system set log_archive_dest_1='LOCATION=/rman/pancake/logship' scope=memory;
System altered.
SQL> recover standby database until cancel;
ORA-00279: change 560224 generated at 05/24/2016 03:25:53 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_49_912576125.dbf
ORA-00280: change 560224 for thread 1 is in sequence #49
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
AUTO
ORA-00279: change 560353 generated at 05/24/2016 03:29:17 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_50_912576125.dbf
ORA-00280: change 560353 for thread 1 is in sequence #50
ORA-00278: log file '/rman/pancake/logship/1_49_912576125.dbf' no longer needed
for this recovery
...
ORA-00279: change 560591 generated at 05/24/2016 03:33:56 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_54_912576125.dbf
ORA-00280: change 560591 for thread 1 is in sequence #54
ORA-00278: log file '/rman/pancake/logship/1_53_912576125.dbf' no longer needed
for this recovery
ORA-00308: cannot open archived log '/rman/pancake/logship/1_54_912576125.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3

最终归档日志回复报告错误、但这是正常的。此错误指示sqlplus正在查找特定日志文件、但未找到该文件。原因很可能是日志文件尚不存在。

增量日志复制和重放

在大多数情况下、不会立即执行迁移。迁移过程可能需要几天甚至几周时间才能完成、这意味着必须将日志持续运送到副本数据库并进行重新显示。这样可以确保在转换到达时传输和回调的数据最少。

可以轻松编写此过程的脚本。例如、可以在原始数据库上计划以下命令、以确保用于日志传送的位置持续更新。

[oracle@jfsc1 pancake]$ cat copylogs.rman
configure channel device type disk format '/rman/pancake/logship/%h_%e_%a.dbf';
backup as copy archivelog from time 'sysdate-2';

[oracle@jfsc1 pancake]$ rman target / cmdfile=copylogs.rman
Recovery Manager: Release 12.1.0.2.0 - Production on Tue May 24 04:36:19 2016
Copyright (c) 1982, 2014, Oracle and/or its affiliates.  All rights reserved.
connected to target database: PANCAKE (DBID=3574534589)
RMAN> configure channel device type disk format '/rman/pancake/logship/%h_%e_%a.dbf';
2> backup as copy archivelog from time 'sysdate-2';
3>
4>
using target database control file instead of recovery catalog
old RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/logship/%h_%e_%a.dbf';
new RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/logship/%h_%e_%a.dbf';
new RMAN configuration parameters are successfully stored
Starting backup at 24-MAY-16
current log archived
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=369 device type=DISK
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=54 RECID=123 STAMP=912659482
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:36:22
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_54_912576125.dbf
continuing other job steps, job failed will not be re-run
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=41 RECID=124 STAMP=912659483
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:36:23
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_41_912576125.dbf
continuing other job steps, job failed will not be re-run
...
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=45 RECID=152 STAMP=912659514
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:36:55
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_45_912576125.dbf
continuing other job steps, job failed will not be re-run
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=47 RECID=153 STAMP=912659515
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:36:57
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_47_912576125.dbf
Recovery Manager complete.

收到日志后、必须对其进行重新显示。前面的示例显示了如何使用sqlplus手动运行 recover database until cancel，可以轻松实现自动化。此处显示的示例使用中所述的脚本 "在备用数据库上重放日志"。该脚本接受一个参数、用于指定需要重放操作的数据库。此过程允许在多数据库迁移工作中使用相同的脚本。

[root@jfsc2 pancake]# ./replaylogs.pl PANCAKE
ORACLE_SID = [oracle] ? The Oracle base has been set to /orabin
SQL*Plus: Release 12.1.0.2.0 Production on Tue May 24 04:47:10 2016
Copyright (c) 1982, 2014, Oracle.  All rights reserved.
Connected to:
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options
SQL> ORA-00279: change 560591 generated at 05/24/2016 03:33:56 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_54_912576125.dbf
ORA-00280: change 560591 for thread 1 is in sequence #54
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
ORA-00279: change 562219 generated at 05/24/2016 04:15:08 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_55_912576125.dbf
ORA-00280: change 562219 for thread 1 is in sequence #55
ORA-00278: log file '/rman/pancake/logship/1_54_912576125.dbf' no longer needed for this recovery
ORA-00279: change 562370 generated at 05/24/2016 04:19:18 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_56_912576125.dbf
ORA-00280: change 562370 for thread 1 is in sequence #56
ORA-00278: log file '/rman/pancake/logship/1_55_912576125.dbf' no longer needed for this recovery
...
ORA-00279: change 563137 generated at 05/24/2016 04:36:20 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_65_912576125.dbf
ORA-00280: change 563137 for thread 1 is in sequence #65
ORA-00278: log file '/rman/pancake/logship/1_64_912576125.dbf' no longer needed for this recovery
ORA-00308: cannot open archived log '/rman/pancake/logship/1_65_912576125.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
SQL> Disconnected from Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options

转换

准备好转换到新环境后、必须执行一次最终同步。使用常规文件系统时、可以轻松确保迁移的数据库与原始数据库100%同步、因为原始重做日志会被复制和重做。使用ASM无法实现此目的。只能轻松地重新复制归档日志。为了确保不会丢失任何数据、必须谨慎地最终关闭原始数据库。

首先、必须将数据库静机、以确保不会进行任何更改。此暂停可能包括禁用计划的操作、关闭侦听器和/或关闭应用程序。
执行此步骤后、大多数数据库配置协议都会创建一个虚拟表、用作关闭标记。

强制进行日志归档、以确保在归档日志中记录虚拟表的创建。为此、请运行以下命令：

SQL> create table cutovercheck as select * from dba_users;
Table created.
SQL> alter system archive log current;
System altered.
SQL> shutdown immediate;
Database closed.
Database dismounted.
ORACLE instance shut down.

要复制最后一个归档日志、请运行以下命令。数据库必须可用、但未打开。

SQL> startup mount;
ORACLE instance started.
Total System Global Area  805306368 bytes
Fixed Size                  2929552 bytes
Variable Size             331353200 bytes
Database Buffers          465567744 bytes
Redo Buffers                5455872 bytes
Database mounted.

要复制归档日志、请运行以下命令：

RMAN> configure channel device type disk format '/rman/pancake/logship/%h_%e_%a.dbf';
2> backup as copy archivelog from time 'sysdate-2';
3>
4>
using target database control file instead of recovery catalog
old RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/logship/%h_%e_%a.dbf';
new RMAN configuration parameters:
CONFIGURE CHANNEL DEVICE TYPE DISK FORMAT   '/rman/pancake/logship/%h_%e_%a.dbf';
new RMAN configuration parameters are successfully stored
Starting backup at 24-MAY-16
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=8 device type=DISK
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=54 RECID=123 STAMP=912659482
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:58:24
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_54_912576125.dbf
continuing other job steps, job failed will not be re-run
...
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=45 RECID=152 STAMP=912659514
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:58:58
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_45_912576125.dbf
continuing other job steps, job failed will not be re-run
channel ORA_DISK_1: starting archived log copy
input archived log thread=1 sequence=47 RECID=153 STAMP=912659515
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 05/24/2016 04:59:00
ORA-19635: input and output file names are identical: /rman/pancake/logship/1_47_912576125.dbf

最后、在新服务器上重放其余归档日志。

[root@jfsc2 pancake]# ./replaylogs.pl PANCAKE
ORACLE_SID = [oracle] ? The Oracle base has been set to /orabin
SQL*Plus: Release 12.1.0.2.0 Production on Tue May 24 05:00:53 2016
Copyright (c) 1982, 2014, Oracle.  All rights reserved.
Connected to:
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options
SQL> ORA-00279: change 563137 generated at 05/24/2016 04:36:20 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_65_912576125.dbf
ORA-00280: change 563137 for thread 1 is in sequence #65
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
ORA-00279: change 563629 generated at 05/24/2016 04:55:20 needed for thread 1
ORA-00289: suggestion : /rman/pancake/logship/1_66_912576125.dbf
ORA-00280: change 563629 for thread 1 is in sequence #66
ORA-00278: log file '/rman/pancake/logship/1_65_912576125.dbf' no longer needed
for this recovery
ORA-00308: cannot open archived log '/rman/pancake/logship/1_66_912576125.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
SQL> Disconnected from Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options

在此阶段、复制所有数据。数据库已准备好从备用数据库转换为活动操作数据库、然后再打开。
```
SQL> alter database activate standby database;
Database altered.
SQL> alter database open;
Database altered.
```

确认是否存在假表、然后将其放下。

SQL> desc cutovercheck
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 USERNAME                                  NOT NULL VARCHAR2(128)
 USER_ID                                   NOT NULL NUMBER
 PASSWORD                                           VARCHAR2(4000)
 ACCOUNT_STATUS                            NOT NULL VARCHAR2(32)
 LOCK_DATE                                          DATE
 EXPIRY_DATE                                        DATE
 DEFAULT_TABLESPACE                        NOT NULL VARCHAR2(30)
 TEMPORARY_TABLESPACE                      NOT NULL VARCHAR2(30)
 CREATED                                   NOT NULL DATE
 PROFILE                                   NOT NULL VARCHAR2(128)
 INITIAL_RSRC_CONSUMER_GROUP                        VARCHAR2(128)
 EXTERNAL_NAME                                      VARCHAR2(4000)
 PASSWORD_VERSIONS                                  VARCHAR2(12)
 EDITIONS_ENABLED                                   VARCHAR2(1)
 AUTHENTICATION_TYPE                                VARCHAR2(8)
 PROXY_ONLY_CONNECT                                 VARCHAR2(1)
 COMMON                                             VARCHAR2(3)
 LAST_LOGIN                                         TIMESTAMP(9) WITH TIME ZONE
 ORACLE_MAINTAINED                                  VARCHAR2(1)
SQL> drop table cutovercheck;
Table dropped.

无中断重做日志迁移

有时、除了重做日志之外、数据库整体组织正确。发生这种情况的原因有很多、其中最常见的原因是与快照有关。SnapManager for Oracle、SnapCenter和NetApp Snap Creator存储管理框架等产品可以近乎即时地恢复数据库、但前提是您还原数据文件卷的状态。如果重做日志与数据文件共享空间、则无法安全地执行还原、因为它会导致重做日志被销毁、这可能意味着数据丢失。因此、必须重新定位重做日志。

此操作步骤非常简单、可以无干扰地执行。

当前重做日志配置

确定重做日志组的数量及其相应的组编号。

SQL> select group#||' '||member from v$logfile;
GROUP#||''||MEMBER
--------------------------------------------------------------------------------
1 /redo0/NTAP/redo01a.log
1 /redo1/NTAP/redo01b.log
2 /redo0/NTAP/redo02a.log
2 /redo1/NTAP/redo02b.log
3 /redo0/NTAP/redo03a.log
3 /redo1/NTAP/redo03b.log
rows selected.

输入重做日志的大小。

SQL> select group#||' '||bytes from v$log;
GROUP#||''||BYTES
--------------------------------------------------------------------------------
1 524288000
2 524288000
3 524288000

创建新日志

对于每个重做日志、创建一个大小和成员数量匹配的新组。

SQL> alter database add logfile ('/newredo0/redo01a.log', '/newredo1/redo01b.log') size 500M;
Database altered.
SQL> alter database add logfile ('/newredo0/redo02a.log', '/newredo1/redo02b.log') size 500M;
Database altered.
SQL> alter database add logfile ('/newredo0/redo03a.log', '/newredo1/redo03b.log') size 500M;
Database altered.
SQL>

验证新配置。

SQL> select group#||' '||member from v$logfile;
GROUP#||''||MEMBER
--------------------------------------------------------------------------------
1 /redo0/NTAP/redo01a.log
1 /redo1/NTAP/redo01b.log
2 /redo0/NTAP/redo02a.log
2 /redo1/NTAP/redo02b.log
3 /redo0/NTAP/redo03a.log
3 /redo1/NTAP/redo03b.log
4 /newredo0/redo01a.log
4 /newredo1/redo01b.log
5 /newredo0/redo02a.log
5 /newredo1/redo02b.log
6 /newredo0/redo03a.log
6 /newredo1/redo03b.log
12 rows selected.

丢弃旧日志

丢弃旧日志(组1、2和3)。

SQL> alter database drop logfile group 1;
Database altered.
SQL> alter database drop logfile group 2;
Database altered.
SQL> alter database drop logfile group 3;
Database altered.

如果遇到错误、导致您无法删除活动日志、请强制切换到下一个日志以释放锁定并强制执行全局检查点。请参见以下此过程的示例。删除位于旧位置的日志文件组2的尝试被拒绝、因为此日志文件中仍有活动数据。

SQL> alter database drop logfile group 2;
alter database drop logfile group 2
*
ERROR at line 1:
ORA-01623: log 2 is current log for instance NTAP (thread 1) - cannot drop
ORA-00312: online log 2 thread 1: '/redo0/NTAP/redo02a.log'
ORA-00312: online log 2 thread 1: '/redo1/NTAP/redo02b.log'

日志归档后加上检查点可用于删除日志文件。

SQL> alter system archive log current;
System altered.
SQL> alter system checkpoint;
System altered.
SQL> alter database drop logfile group 2;
Database altered.

然后从文件系统中删除日志。执行此过程时应格外小心。

通过日志传送实现Oracle数据库迁移

Creating your file...

日志传送-文件系统到文件系统

创建数据库备份

environment

还原到新环境

创建控制文件创建模板

备份参数文件

创建oratab条目

准备目录结构

参数文件更新

重新创建控制文件

初始日志复制

初始日志重放

增量日志复制和重放

转换

日志传送- ASM到文件系统

创建数据库备份

备份控制文件

备份参数文件

ASM文件重命名脚本

准备目录结构

创建oratab条目

参数更新

启动非挂载

复制数据库

初始日志复制

初始日志重放

增量日志复制和重放

转换

无中断重做日志迁移

当前重做日志配置

创建新日志

丢弃旧日志