完整的oraclerman備份恢復(fù)的例子

1、完整的 oracle rman 備份恢復(fù)的例子作者： zhumy1、 建 rman 庫作為 repository$more createrman_db1.shset echo onspool makedb1.logcreate database rmandatafile /export/home/oracle/oradata/rman_data/system.dbf size 50m autoextendon next 640Klogfile /export/home/oracle/oradata/rman_data/redo0101.log SIZE 10M,/export/home/ora

2、cle/oradata/rman_data/redo0201.log SIZE 10Mmaxdatafiles 30maxinstances 8maxlogfiles 64character set US7ASCIInational character set US7ASCII;disconnectspool offexit/export/home/oracle/8.1.6/rdbms/admin/catalog.sql;REM * ALTER SYSTEM TABLESPACE *ALTER TABLESPACE SYSTEMDEFAULT STORAGE ( INITIAL 64K NEX

3、T 64K MINEXTENTS 1 MAXEXTENTS UNLIMITEDPCTINCREASE 50);ALTER TABLESPACE SYSTEMMINIMUM EXTENT 64K;REM * TABLESPACE FOR ROLLBACK *CREATE TABLESPACE RBS DATAFILE /export/home/oracle/oradata/rman_data/rbs.dbf size 50mAUTOEXTEND ON NEXT 512KMINIMUM EXTENT 512KDEFAULT STORAGE ( INITIAL 512K NEXT 512K MINE

4、XTENTS 10 MAXEXTENTS UNLIMITED PC TINCREASE 0 );REM * TABLESPACE FOR TEMPORARY *CREATE TABLESPACE TEMP DATAFILE /export/home/oracle/oradata/rman_data/temp.dbf size 50mAUTOEXTEND ON NEXT 64KMINIMUM EXTENT 64KDEFAULT STORAGE ( INITIAL 64K NEXT 64K MINEXTENTS 1 MAXEXTENTS UNLIMITEDPCTINCREASE 0) TEMPOR

5、ARY;REM * Creating four rollback segments *CREATE PUBLIC ROLLBACK SEGMENT RBS_0 TABLESPACE RBSSTORAGE ( OPTIMAL 64000K );ALTER ROLLBACK SEGMENT RBS_0 ONLINE;REM*SYS and SYSTEM users*alter user sys temporary tablespace TEMP;alter user system temporary tablespace TEMP;disconnect spool offexit$more cre

6、aterman_db3.shspool crdb3.log/export/home/oracle/8.1.6/rdbms/admin/catproc.sql/export/home/oracle/8.1.6/rdbms/admin/caths.sql/export/home/oracle/8.1.6/rdbms/admin/otrcsvr.sqlconnect system/manager/export/home/oracle/8.1.6/sqlplus/admin/pupbld.sqldisconnectspool offexit2 、建 repository 存放的表空間和rman 用戶$

7、more createrman_db4.shconnect internalcreate tablespace rman_tsdatafile /export/home/oracle/oradata/rman_data/rman_ts.dbfsize 20M default storage (initial 100K next 100K pctincrease 0);create user rman_hainan identified by rman_hainantemporary tablespace TEMPdefault tablespace rman_ts quota unlimite

8、d onrman_ts;grant recovery_catalog_owner to rman_hainan;grant connect ,resource to rman_hainan;3 、建 catalog ，注冊目標(biāo)數(shù)據(jù)庫$more createrman_db5.shrman catalog rman_hainan/rman_hainanrman msglog=rman.logcreate catalog ;exit;rman target sys/oracledb1connect catalog rman_hainan/rman_hainanrmanregister databas

9、e;exit;4、可以開始做備份了。5、做全備$more rmanshell. /export/home/oracle/.profilerman rcvcat rman_hainan/rman_hainanrman target / cmdfile /export/home/oracle/backup_level0.rcv log /export/home/oracle/backup.log$more backup_level0.rcvresync catalog;run allocate channel t1 type disk;backupincremental level 0skip i

10、naccessibletag hot_db_bk_level0filesperset 3format /export/home/oracle/bk_%s_%p_%t.bk(database);sql alter system archive log current;backup filesperset 10format /export/home/oracle/a1_%s_%p_%t.ac(archivelog all delete input);backupformat /export/home/oracle/df_t%t_s%s_p%p.ctcurrent controlfile ;6、做增

11、備$more rmanshell1rman rcvcat rman_hainan/rman_hainanrman target / cmdfile backup_level1.rcv log backup.log$more backup_level1.rcvresync catalog;run allocate channel t1 type disk;backupincremental level 1skip inaccessibletag hot_db_bk_level1filesperset 3format bk_%s_%p_%t.bk1(database);sql alter syst

12、em archive log current;backupfilesperset 10format a1_%s_%p_%t.ac1(archivelog all delete input);backup current controlfile;1、 刪除舊的全備$rman rcvcat rman_hainan/rman_hainanrman target /Recovery Manager: Release .0 - ProductionRMAN-06005: connected to target database: TEST (DBID=1692992254)RMAN-060

13、08: connected to recovery catalog databaseRMAN list backupset;RMAN-03022: compiling command: listList of Backup SetsKey Recid Stamp LV Set Stamp Set Count Completion Time38 145 399987408 0 399987406 153 11-JUN-00根據(jù) key 來刪除舊的備份。RMAN allocate channel for maintenance type disk;RMAN change backupset 169

14、 delete; THIS IS THE COMMAND TO REMOVE THE ENTRY& OS FILE做完后可以看到 list backupset 和操作系統(tǒng)的文件都沒有了。2 、 恢復(fù)( 1) 將數(shù)據(jù)庫啟動到 nomount 狀態(tài)：$svrmgrlOracle Server Manager Release .0 - ProductionCopyright (c) 1997, 1999, Oracle Corporation. All Rights Reserved.Oracle8i Enterprise Edition Release .0 - 64b

15、it ProductionWith the Partitioning optionJServer Release .0 - ProductionSVRMGR connect internalConnected.SVRMGR startup nomount;ORACLE instance started.Total System Global Area 339275684 bytesFixed Size 94116 bytesVariable Size 318685184 bytesDatabase Buffers 16384000 bytesRedo Buffers 411238

16、4 bytesSVRMGR exitServer Manager complete.( 2 ) 恢復(fù)控制文件：$rman rcvcat rman_hainan/rman_hainanrman target /Recovery Manager: Release .0 - ProductionRMAN-06006: connected to target database: test (not mounted)RMAN-06008: connected to recovery catalog databaseRMAN run 2 allocate channel d1 type di

17、sk;3 restore controlfile;4 release channel d1;5 （ 3 ） 恢復(fù)數(shù)據(jù)文件RMAN run 2 allocate channel d1 type disk;3 sql alter database mount;4 restore datafile 1;5 restore datafile 2;6 restore datafile 3;7 restore datafile 4;8 release channel d1;9 （ 4 ） 恢復(fù)日志文件RMAN run 2 set archivelog destination to /export/home

18、/oracle/admin/test/arch;3 allocate channel d1 type disk;4 restore archivelog all;5 release channel d1;6 會把所有的日志文件恢復(fù)。( 5 ) 根據(jù)日志做recover$svrmgrlOracle Server Manager Release .0 - ProductionCopyright (c) 1997, 1999, Oracle Corporation. All Rights Reserved.Oracle8i Enterprise Edition Release 8.1.

19、6.0.0 - 64bit ProductionWith the Partitioning optionJServer Release .0 - ProductionSVRMGR connect internalConnected.SVRMGR recover database using backup controlfile until cancel;ORA-00279: change 51054 generated at 06/11/2000 11:38:37 needed for thread 1ORA-00289: suggestion : /export/home/or

20、acle/admin/test/arch/arch_1_3.arcORA-00280: change 51054 for thread 1 is in sequence #3Specify log: =suggested | filename | AUTO | CANCELLog applied.ORA-00279: change 51058 generated at 06/11/2000 11:38:44 needed for thread 1ORA-00289: suggestion : /export/home/oracle/admin/test/arch/arch_1_4.arcORA

21、-00280: change 51058 for thread 1 is in sequence #4ORA-00278: log file /export/home/oracle/admin/test/arch/arch_1_3.arc no longerneeded for this recoverySpecify log: =suggested | filename | AUTO | CANCELLog applied.ORA-00279: change 51074 generated at 06/11/2000 11:40:20 needed for thread 1ORA-00289

22、: suggestion : /export/home/oracle/admin/test/arch/arch_1_5.arcORA-00280: change 51074 for thread 1 is in sequence #5ORA-00278: log file /export/home/oracle/admin/test/arch/arch_1_4.arc no longerneeded for this recoverySpecify log: =suggested | filename | AUTO | CANCELcancelMedia recovery cancelled.SVRMGR alter database open resetlogs;Statement processed