使用RMAN将RAC+ASM复制到单实例+ASM上

本次试验将部署到ASM上的双节点RAC复制到单实例的ASM上。

该试验也是出自于一个项目的需求，在虚拟机上演练操作特此记录。

之前已经做过多次RMAN复制数据库，这次试验的特点是从RAC复制到单实例。需要注意的以下内容：

1，spfile参数问题
RAC下的spfile参数中记录了很多和集群相关的信息，在复制时需要修改。比如
*.cluster_database=true
PROD2.instance_number=2
PROD1.instance_number=1
*.remote_listener='cluster-scan:1521'
PROD2.thread=2
PROD1.thread=1
PROD2.undo_tablespace='UNDOTBS2'
PROD1.undo_tablespace=‘UNDOTBS1’
和RAC相关的信息可以选择注释掉，或者删除。

2，undo表空间问题
双节点RAC的备份集中会产生2个undo表空间及其数据文件，而单实例数据库只需要一个undo表空间。使用RAC的备份集在单实例数据库上恢复后，可以手工删除多余的那个undo表空间及其数据文件，通常是undotbs2

3，redo问题
双节点RAC的控制文件的备份中记录的是两个实例的redo信息，恢复完成后在open resetlogs打开数据库后回创建出thread2的online redo log，也就是节点2的联机日志。同undotbs2的处理原则一样，删除之。


1，RAC 端rman全备数据库到/backup目录下。

[oracle@node1 ~]$ rman target /

Recovery Manager: Release 11.2.0.4.0 - Production on Mon Sep 29 03:15:49 2014

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.

connected to target database: PROD (DBID=271163854)

RMAN> run{
2> allocate channel c1 type disk;
3> allocate channel c2 type disk;
4> backup tag 'full' format '/backup/full_%U.bak'
5> database include current controlfile;
6> sql 'alter system archive log current';
7> backup tag 'arch' format '/backup/arch_%U.arc'
8> archivelog all;
9> release channel c1;
10> release channel c2;
11> }

RMAN> list backup;

List of Backup Sets
===================

BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
43      Full    522.99M    DISK        00:00:04     2014-09-29 03:05:29
        BP Key: 43   Status: AVAILABLE  Compressed: NO  Tag: FULL
        Piece Name: /backup/full_1bpjmdbl_1_1.bak
  List of Datafiles in backup set 43
  File LV Type Ckp SCN    Ckp Time            Name
  ---- -- ---- ---------- ------------------- ----
  2       Full 1298717    2014-09-29 03:05:25 +DATA/prod/datafile/sysaux.257.859325451
  4       Full 1298717    2014-09-29 03:05:25 +DATA/prod/datafile/users.259.859325451
  5       Full 1298717    2014-09-29 03:05:25 +DATA/prod/datafile/example.264.859325525
  6       Full 1298717    2014-09-29 03:05:25 +DATA/prod/datafile/undotbs2.265.859325695

BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
44      Full    649.16M    DISK        00:00:04     2014-09-29 03:05:29
        BP Key: 44   Status: AVAILABLE  Compressed: NO  Tag: FULL
        Piece Name: /backup/full_1cpjmdbl_1_1.bak
  List of Datafiles in backup set 44
  File LV Type Ckp SCN    Ckp Time            Name
  ---- -- ---- ---------- ------------------- ----
  1       Full 1298719    2014-09-29 03:05:25 +DATA/prod/datafile/system.256.859325451
  3       Full 1298719    2014-09-29 03:05:25 +DATA/prod/datafile/undotbs1.258.859325451
  7       Full 1298719    2014-09-29 03:05:25 +DATA/prod/datafile/goldengate.269.859338811

BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
45      Full    80.00K     DISK        00:00:00     2014-09-29 03:05:32
        BP Key: 45   Status: AVAILABLE  Compressed: NO  Tag: FULL
        Piece Name: /backup/ full_1epjmdbs_1_1.bak           << 参数文件所在备份片，后面恢复时候需要用到

   SPFILE  Included: Modification time: 2014-09-29 02:02:18  
  SPFILE db_unique_name: PROD

BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
46      Full    17.80M     DISK        00:00:01     2014-09-29 03:05:33
        BP Key: 46   Status: AVAILABLE  Compressed: NO  Tag: FULL
        Piece Name: /backup/ full_1dpjmdbs_1_1.bak             << 控制文件所在备份片，后面恢复时候需要用到

   Control File  Included: Ckp SCN: 1298731      Ckp time: 2014-09-29 03:05:32

BS Key  Size       Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
47      53.63M     DISK        00:00:00     2014-09-29 03:05:44
        BP Key: 47   Status: AVAILABLE  Compressed: NO  Tag: ARCH
        Piece Name: /backup/arch_1gpjmdc8_1_1.arc

  List of Archived Logs in backup set 47
  Thrd Seq     Low SCN    Low Time            Next SCN   Next Time
  ---- ------- ---------- ------------------- ---------- ---------
  1    28      1260638    2014-09-28 23:53:08 1260655    2014-09-28 23:53:11
  1    29      1260655    2014-09-28 23:53:11 1298750    2014-09-29 03:05:34
  1    30      1298750    2014-09-29 03:05:34 1298772    2014-09-29 03:05:40
  2    24      1260645    2014-09-28 23:53:19 1260660    2014-09-28 23:53:22
  2    25      1260660    2014-09-28 23:53:22 1298755    2014-09-29 03:05:42
  2    26      1298755    2014-09-29 03:05:42 1298767    2014-09-29 03:05:45

BS Key  Size       Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
48      62.20M     DISK        00:00:00     2014-09-29 03:05:44
        BP Key: 48   Status: AVAILABLE  Compressed: NO  Tag: ARCH
        Piece Name: /backup/arch_1fpjmdc8_1_1.arc

  List of Archived Logs in backup set 48
  Thrd Seq     Low SCN    Low Time            Next SCN   Next Time
  ---- ------- ---------- ------------------- ---------- ---------
  1    25      1211968    2014-09-28 19:28:25 1221060    2014-09-28 20:26:23
  1    26      1221060    2014-09-28 20:26:23 1221084    2014-09-28 20:26:29
  1    27      1221084    2014-09-28 20:26:29 1260638    2014-09-28 23:53:08
  2    21      1211964    2014-09-28 19:28:23 1221067    2014-09-28 20:26:25
  2    22      1221067    2014-09-28 20:26:25 1221089    2014-09-28 20:26:31
  2    23      1221089    2014-09-28 20:26:31 1260645    2014-09-28 23:53:19

2，RAC端使用scp传输备份集到单实例端/home/oracle目录下
[oracle@node1 ~]$ scp /backup/* oracle@172.16.228.8:/home/oracle

单实例端查看接收到的备份集
[oracle@single ~]$ ls -l /home/oracle
total 1338624
-rw-r----- 1 oracle oinstall  65219584 Oct  8 21:39 arch_1fpjmdc8_1_1.arc
-rw-r----- 1 oracle oinstall  56237568 Oct  8 21:39 arch_1gpjmdc8_1_1.arc
-rw-r----- 1 oracle oinstall 548405248 Oct  8 21:39 full_1bpjmdbl_1_1.bak
-rw-r----- 1 oracle oinstall 680697856 Oct  8 21:40 full_1cpjmdbl_1_1.bak
-rw-r----- 1 oracle oinstall  18677760 Oct  8 21:40 full_1dpjmdbs_1_1.bak
-rw-r----- 1 oracle oinstall     98304 Oct  8 21:40 full_1epjmdbs_1_1.bak

3，单实例端使用RMAN从备份集中恢复pfile到/home/oracle/racpfile.ora
$ rman target /

RMAN> startup nomount;

RMAN> restore spfile to pfile '/home/oracle/racpfile.ora' from '/home/oracle/full_1epjmdbs_1_1.bak';

4， 单实例端 修改pfile文件，将集群相关的信息全部注释掉
[oracle@single ~]$ cat racpfile.ora
##PROD1.__db_cache_size=146800640
##PROD2.__db_cache_size=176160768
##PROD1.__java_pool_size=4194304
##PROD2.__java_pool_size=4194304
##PROD1.__large_pool_size=8388608
##PROD2.__large_pool_size=8388608
##PROD1.__oracle_base='/u01/app/oracle'#ORACLE_BASE set from environment
##PROD1.__pga_aggregate_target=289406976
##PROD2.__pga_aggregate_target=293601280
##PROD1.__sga_target=385875968
##PROD2.__sga_target=381681664
##PROD1.__shared_io_pool_size=0
##PROD2.__shared_io_pool_size=0
##PROD1.__shared_pool_size=218103808
##PROD2.__shared_pool_size=184549376
##PROD1.__streams_pool_size=0
##PROD2.__streams_pool_size=0
##*.audit_file_dest='/u01/app/oracle/admin/PROD/adump'
*.audit_file_dest='/u01/admin/PROD/adump'
*.audit_trail='db'
##*.cluster_database=true   << 集群
*.compatible='11.2.0.4.0'
##*.control_files='+DATA/prod/controlfile/current.260.859325519','+ARCH/prod/controlfile/current.256.859325519'
*.control_files='+DATA/PROD/controlfile/control01.ctl','+FRA/PROD/controlfile/control02.ctl'
*.db_block_size=8192
*.db_create_file_dest='+DATA'
*.db_domain=''
*.db_name='PROD'
##*.db_recovery_file_dest='+ARCH'
*.db_recovery_file_dest='+FRA'
*.db_recovery_file_dest_size=4621074432
##*.diagnostic_dest='/u01/app/oracle'
*.diagnostic_dest='/u01/'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=PRODXDB)'
*.enable_goldengate_replication=TRUE
##PROD2.instance_number=2
##PROD1.instance_number=1
##*.memory_target=675282944
*.memory_target=583008256
*.open_cursors=300
*.processes=150
##*.remote_listener='cluster-scan:1521'   << scan IP
*.remote_login_passwordfile='exclusive'
##PROD2.thread=2
##PROD1.thread=1
##PROD2.undo_tablespace='UNDOTBS2'
##PROD1.undo_tablespace='UNDOTBS1'
*.undo_tablespace='UNDOTBS1'                 << 此处只保留unodtbs1的参数，在后面的恢复中还是会将undotbs2给恢复出来。

5，单实例端SQLPLUS使用修改后的pfile文件启动实例到nomount状态
SYS@PROD >startup nomount pfile=/home/oracle/racpfile.ora;

6，单实例端创建spfile到ASM磁盘组中的+DATA/PROD
SYS@PROD >create spfile='+DATA/PROD/spfilePROD.ora' from memory;

7，单实例端创建pfile，指引spfile的文件位置
$ vi $ORACLE_HOME/dbs/initPROD.ora

spfile='+DATA/PROD/spfilePROD.ora'

8，单实例端使用RMAN启动到nomount状态
RMAN> startup force nomount;

RMAN> set DBID=271163854

9，单实例端使用RAMN从备份集中恢复控制文件
RMAN> restore controlfile from '/home/oracle/full_1dpjmdbs_1_1.bak’;

10，单实例端mount数据库
RMAN> mount database;

11，单实例端注册备份集的路径
RMAN> catalog start with '/home/oracle';

12，单实例端列出控制文件中记录的数据文件
RMAN> report schema;

RMAN-06139: WARNING: control file is not current for REPORT SCHEMA
Report of database schema for database with db_unique_name PROD

List of Permanent Datafiles
===========================
File Size(MB) Tablespace           RB segs Datafile Name
---- -------- -------------------- ------- ------------------------
1    0        SYSTEM               ***     +DATA/prod/datafile/system.256.859325451
2    0        SYSAUX               ***     +DATA/prod/datafile/sysaux.257.859325451
3    0        UNDOTBS1             ***     +DATA/prod/datafile/undotbs1.258.859325451
4    0        USERS                ***     +DATA/prod/datafile/users.259.859325451
5    0        EXAMPLE              ***     +DATA/prod/datafile/example.264.859325525
6    0        UNDOTBS2             ***     +DATA/prod/datafile/undotbs2.265.859325695
7    0        GOLDENGATE           ***     +DATA/prod/datafile/goldengate.269.859338811

List of Temporary Files
=======================
File Size(MB) Tablespace           Maxsize(MB) Tempfile Name
---- -------- -------------------- ----------- --------------------
1    20       TEMP                 32767       +DATA/prod/tempfile/temp.263.859325523

13，单实例端根据列出的数据文件采用set newname的方式来修改文件路径。
RMAN> run{
2> set newname for datafile 1 to '+DATA';
3> set newname for datafile 2 to '+DATA';
4> set newname for datafile 3 to '+DATA';
5> set newname for datafile 4 to '+DATA';
6> set newname for datafile 5 to '+DATA';
7> set newname for datafile 6 to '+DATA';
8> set newname for datafile 7 to '+DATA';
9> set newname for tempfile 1 to '+DATA’;   <
10> restore database;
11> switch datafile all;
12> switch tempfile all;
13> recover database;
14> }

14，单实例端resetlogs方式打开数据库库
RMAN> alter database open resetlogs;


15，单实例端查看redo log的分布情况，注意thread=2是RAC中第二个节点上的，单实例中不需要刻意删除掉。
SYS@PROD >SELECT v$logfile.member, v$logfile.group#,thread#, v$log.status,v$log.ARCHIVED, v$log.bytes/1024/1024 ,v$log.thread#
  2  FROM v$log, v$logfile
  3  WHERE v$log.group# = v$logfile.group#
  4  ORDER BY v$log.thread#,v$logfile.group#;

MEMBER                                                 GROUP#    THREAD# STATUS           ARC V$LOG.BYTES/1024/1024    THREAD#
-------------------------------------------------- ---------- ---------- ---------------- --- --------------------- ----------
+FRA/prod/onlinelog/group_1.308.860453423                   1          1 CURRENT          NO                     50          1
+DATA/prod/onlinelog/group_1.286.860453423                  1          1 CURRENT          NO                     50          1
+FRA/prod/onlinelog/group_2.261.860453423                   2          1 UNUSED           YES                    50          1
+DATA/prod/onlinelog/group_2.263.860453423                  2          1 UNUSED           YES                    50          1
+DATA/prod/onlinelog/group_3.295.860453423                  3          2 INACTIVE         YES                    50          2
+FRA/prod/onlinelog/group_3.260.860453423                   3          2 INACTIVE         YES                    50          2
+DATA/prod/onlinelog/group_4.258.860453423                  4          2 UNUSED           YES                    50          2
+FRA/prod/onlinelog/group_4.307.860453423                   4          2 UNUSED           YES                    50          2

16，禁用thread 2
SYS@PROD >alter database disable thread 2;

17，删除thread=2的日志组group3和group4
SYS@PROD >alter database drop logfile group 3;

SYS@PROD >alter database drop logfile group 4;

删除thread=2的日志组后的redo log分布
SYS@PROD >SELECT v$logfile.member, v$logfile.group#,thread#, v$log.status,v$log.ARCHIVED, v$log.bytes/1024/1024 ,v$log.thread#
  2  FROM v$log, v$logfile
  3  WHERE v$log.group# = v$logfile.group#
  4  ORDER BY v$log.thread#,v$logfile.group#;

MEMBER                                                 GROUP#    THREAD# STATUS           ARC V$LOG.BYTES/1024/1024    THREAD#
-------------------------------------------------- ---------- ---------- ---------------- --- --------------------- ----------
+DATA/prod/onlinelog/group_1.286.860453423                  1          1 CURRENT          NO                     50          1
+FRA/prod/onlinelog/group_1.308.860453423                   1          1 CURRENT          NO                     50          1
+DATA/prod/onlinelog/group_2.263.860453423                  2          1 UNUSED           YES                    50          1
+FRA/prod/onlinelog/group_2.261.860453423                   2          1 UNUSED           YES                    50          1

18，同样思路，删除不属于单实例的undotbs2表空间及其数据文件
SYS@PROD >col name for a50
SYS@PROD >select name from v$datafile;

NAME
--------------------------------------------------
+DATA/prod/datafile/system.296.860453223
+DATA/prod/datafile/sysaux.262.860453229
+DATA/prod/datafile/undotbs1.273.860453223
+DATA/prod/datafile/users.261.860453229
+DATA/prod/datafile/example.293.860453229
+DATA/prod/datafile/undotbs2.285.860453229      << 多余的undotbs2
+DATA/prod/datafile/goldengate.259.860453223

7 rows selected.

SYS@PROD >select name from v$tablespace where name like 'UNDO%';

NAME
--------------------------------------------------
UNDOTBS1
UNDOTBS2

SYS@PROD >drop tablespace undotbs2 including contents and datafiles;

此次试验结束

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/29047826/viewspace-1297860/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/29047826/viewspace-1297860/