1. 連接RMAN

SQL> create user rman identified by jansondors;
SQL> grant resource,connect,dba to rman;
$ rman target rman/jansondors

2. 配置RMAN參數(shù)

RMAN> show all;

using target database control file instead of recovery catalog
RMAN configuration parameters for database with db_unique_name ORACENT are:
CONFIGURE RETENTION POLICY TO REDUNDANCY 1; # 保留備份副本的數(shù)量
CONFIGURE BACKUP OPTIMIZATION OFF; # 配置備份優(yōu)化，如果已經(jīng)備份了某個文件的相同版本，則不會再備份該文件
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # 備份的數(shù)據(jù)文件保留至服務器磁盤上
CONFIGURE CONTROLFILE AUTOBACKUP OFF; # 配置是否啟用控制文件的自動備份
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '%F'; # 配置控制文件自動備份的格式
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET; # 備份并行度，通道數(shù)量越多，任務執(zhí)行時間越短；備份文件類型為備份集
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
CONFIGURE COMPRESSION ALGORITHM 'BASIC' AS OF RELEASE 'DEFAULT' OPTIMIZE FOR LOAD TRUE ; # default
CONFIGURE ARCHIVELOG DELETION POLICY TO NONE; # default
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/data/oracle/product/11.2.0/db_1/dbs/snapcf_oracent.f'; # default

RMAN> configure default device type to sbt;
RMAN> configure default device type to disk；
RMAN> configure backup optimization on;
RMAN> configure controlfile autobackup on;

-> 備份路徑遵循的原則
statement format ->configure format ->FLASH_RECOVER_ARER ->$ORACLE_HOME/dbs

3. 配置RMAN脫機備份

$ sqlplus rman/jansondors as sysdba
SQL> select file_type,percent_space_used,percent_space_reclaimable,number_of_files from v$flash_recovery_area_usage;
SQL> shutdown immediate;
SQL> startup mount;
SQL> alter system set db_recovery_file_dest_size=10g;  
SQL> show parameter db_recovery_file_dest;

RMAN> backup database;
RMAN> backup as compressed backupset database;
RMAN> sql 'alter database open';

4. 配置RMAN聯(lián)機備份

• 開啟歸檔

$ sqlplus rman/jansondors as sysdba
SQL> shutdown immediate;
SQL> startup mount;
SQL> alter database archivelog/noarchivelog;
SQL> alter database open;
SQL> archive log list;

• 備份數(shù)據(jù)庫

RMAN> backup as compressed backupset database plus archivelog delete all input;

RMAN> run{
          allocate channel ch1 device type disk format '/home/oracle/Downloads/backup_ctl_file/ch1_%U'；
          allocate channel ch2 device type disk format '/home/oracle/Downloads/backup_ctl_file/ch2_%U'；
          backup as backupset
          (datafile 1,4 channel ch1)
          (datafile 2,3,5 channel ch2);
          sql 'alter system archive log current';}
          
RMAN> run{
          allocate channel ch1 device type disk；
          allocate channel ch2 device type disk；
          backup as backupset format '/home/oracle/Downloads/backup_ctl_file/%U'
          (datafile 1,4 channel ch1)
          (datafile 2,3,5 channel ch2);
          sql 'alter system archive log current';}

• 備份表空間

RMAN> backup tablespace users;
RMAN> backup as compressed backupset tablespace users;

• 備份數(shù)據(jù)文件

RMAN>  backup as backupset datafile 1 format '/home/oracle/Downloads/backup_ctl_file/datafile_1_%U';

• 備份控制文件

->手工備份
RMAN> backup current controlfile format '/home/oracle/Downloads/backup_ctl_file_%U.dbf';
RMAN> backup current controlfile ;

->自動備份
RMAN> configure controlfile autobackup format for device type disk to '/home/oracle/Downloads/backup_ctl_file/%F';
RMAN> configure controlfile autobackup on ;

• 備份壞塊處理

默認情況下，RMAN會檢查數(shù)據(jù)庫是否發(fā)生物理損壞，不會檢查邏輯損壞
->關閉物理損壞檢查
RMAN> backup nochecksum tablespace users tag='weekly backup';
->啟用邏輯損壞檢查
RMAN> backup check logical tablespace users;
->設置maxcorrupt,當壞塊個數(shù)超過指定數(shù)量時，backup進程會停止
RMAN-> run{
           set maxcorrupt for datafile 2,4 to 10;
           backup database;}

5. 配置RMAN增量備份

• Level 0增量備份

RMAN> backup incremental level 0 database;

• Level 1增量備份

RMAN> backup incremental level 1 database;

• 快速增量備份

SQL> shutdown immediate;
SQL> startup mount;

SQL> select filename,status,bytes from v$block_change_tracking;
SQL> alter database enable block change tracking using file '/data/oracle/product/11.2.0/oradata/chtrack.log';
SQL> alter database disable block change tracking;
SQL> alter database rename file '/data/oracle/product/11.2.0/oradata/chtrack.log' to '/data/oracle/product/11.2.0/oradata/test/chtrack.log';

• 應用增量備份

# 將增量備份添加到鏡像副本上

RMAN> run{
          backup incremental level 1 for recover of copy with tag 'incr_copy_backup' database;
          recover copy of database with tag 'incr_copy_backup';}

6. 配置恢復目錄

SQL> create tablespace rcat_tbs datafile '/data/oracle/oradata/ORACENT/rcat_tbs01.dbf' size 100m;
SQL> create user rcat_owner identified by jansondors default tablespace rcat_tbs temporary tablespace temp;
SQL> grant recovery_catalog_owner to rcat_owner;
SQL> grant connect,resource to rcat_owner;

RMAN> rman catalog rman/jansondors target systemn/jansondors@targetOrcl

7. RMAN腳本

create script rman_backup{
 sql 'alter system checkpoint';
 backup database format '/home/oracle/Downloads/offline_backup/backup_$u.dbf';
 backup current controlfile format '/home/oracle/Downloads/offline_backup/backup_ctl_$u.dbf';
}

8. 非歸檔完全恢復

聯(lián)機重做日志是循環(huán)使用的，一個日志寫滿之后會切換到下一個，新的循環(huán)會覆蓋掉部分變化的數(shù)據(jù)，非歸檔恢復是一種不完全恢復

• 數(shù)據(jù)文件、控制文件以及重做日志文件全部丟失

# 數(shù)據(jù)庫處于非歸檔模式
$ sqlplus rman/jansondors as sysdba
SQL> shutdown immediate;
SQL> startup mount;
SQL> alter database noarchivelog;
SQL> alter database open;
SQL> archive log list;

# 備份文件存儲目錄為默認的快閃恢復區(qū)
RMAN> configure controlfile autobackup format for device type disk clear;
RMAN> show all;

# 開啟控制文件自動備份
RMAN> configure controlfile autobackup on;
RMAN> show all;

# 控制文件、數(shù)據(jù)文件、重做日志默認位置
SQL> show parameter control_files;
SQL> col name for a30
SQL> select file#,name,status from v$datafile;
SQL> select group#, status, member from v$logfile;

-> 1. 數(shù)據(jù)庫開啟掛載模式
SQL> startup mount;

-> 2. 新增測試數(shù)據(jù)
SQL> alter database open;
SQL> create table test123 as select * from dba_segments;

-> 3. 模擬文件丟失
SQL> shutdown immediate;
$ pwd
$ ls    
control01.ctl  redo02.log  sysaux01.dbf  temp01.dbf     users01.dbf
redo01.log     redo03.log  system01.dbf  undotbs01.dbf
$ rm -rf *.*

-> 4. 恢復數(shù)據(jù)
SQL> startup
ORACLE instance started.
Total System Global Area  759943168 bytes
Fixed Size          2217224 bytes
Variable Size         503319288 bytes
Database Buffers      251658240 bytes
Redo Buffers            2748416 bytes
ORA-00205: error in identifying control file, check alert log for more info
RMAN> restore controlfile from '/data/oracle/flash_recovery_area/ORACENT/autobackup/2017_05_17/o1_mf_s_944236355_dkr0s40f_.bkp';
SQL> alter database mount;
Database altered.
RMAN> restore database;     
RMAN> recover database noredo;
SQL> alter database open resetlogs;
SQL> select group#,sequence#,status from v$log;
SQL> select count(*) from test123;

• 數(shù)據(jù)文件丟失

-> 1. 模擬USERS表空間丟失
SQL> shutdown immediate
$ pwd
/data/oracle/oracent
$ ls
control01.ctl  redo01.log  redo02.log  redo03.log  sysaux01.dbf  system01.dbf  temp01.dbf  undotbs01.dbf  users01.dbf
$ rm -rf users01.dbf
-> 2. RMAN恢復
SQL> select file#,name from v$datafile;
RMAN> restore datafile 4;
RMAN> recover datafile 4;
#重做日志被覆蓋
RMAN> recover datafile 4 until cancel;

• 重做日志文件丟失

-> 1. 模擬文件丟失
SQL> shutdown immediate;
$ pwd
/data/oracle/oracent
$ ls
control01.ctl  redo01.log  redo02.log  redo03.log  sysaux01.dbf  system01.dbf  temp01.dbf  undotbs01.dbf  users01.dbf
$ rm -rf *.dbf
$ rm -rf *.log
-> 2. 恢復
RMAN> restore database;
RMAN> recover database until cancel;
SQL> alter database open;
SQL> alter database open resetlogs;

• 遷移數(shù)據(jù)文件

SQL> start mount
RMAN> run{
    set newname for datafile
    'A/system01.dbf' to 'B/system01.dbf';
    set newname for datafile
    'A/users01.dbf' to 'B/users01.dbf';
    restore database from tag=TAG20170517T131491;
    switch datafile all;
}

9. 歸檔完全恢復

數(shù)據(jù)庫一直處于歸檔模式下，且歸檔文件和重做日志文件損壞的情況下，可以在聯(lián)機狀態(tài)下恢復數(shù)據(jù)庫文件

• 非系統(tǒng)表空間損壞

-> 1. 模擬環(huán)境
$ pwd
/data/oracle/oracent
$ ls
control01.ctl  redo01.log  redo02.log  redo03.log  sysaux01.dbf  system01.dbf  temp01.dbf  undotbs01.dbf  users01.dbf
$ rm -rf users01.dbf

-> 2. 恢復
SQL> alter database datafile 4 offline;
SQL> alter database open;
RMAN> restore datafile 4;
RMAN> recover datafile 4;
SQL> alter database datafile 4 online;
SQL> select file_name,tablespace_name,status,online_status from dba_data_files;
RMAN> run{
    sql 'alter database datafile 4 offline';
    restore tablespace users;
    recover tablespace users;
    sql 'alter database datafile 4 online';
}

• 系統(tǒng)表空間損壞

# system表空間損壞時，數(shù)據(jù)庫無法啟動；需要把數(shù)據(jù)庫啟動至mount狀態(tài)，使用RMAN進行數(shù)據(jù)庫恢復工作
SQL> startup mount
RMAN> run{
    sql 'alter database datafile 1 offline';
    restore datafile 1 ;
    recover datafile 1;
    sql 'alter database datafile 1 online';
}
SQL> alter database open;

• 所有數(shù)據(jù)文件丟失

# 在所有數(shù)據(jù)文件丟失，但控制文件和重做日志文件都完好的情況下，可以通過RMAN進行數(shù)據(jù)庫恢復工作
RMAN> run{
    restore database;
    recover database;
    sql 'alter database open';
}

10. RMAN恢復數(shù)據(jù)塊

-> 1. 備份整個數(shù)據(jù)庫
RMAN> backup database plus archivelog;

-> 2. 模擬環(huán)境
SQL> shutdown immediate;
# 人為修改dbf文件
# 此時startup指令無效，提示報錯

-> 3. 恢復
RMAN> backup validate datafile 6;
SQL> select * from v$database_block_corruption;
RMAN> blockrecover datafile 6 block #BLOCK_NUM# from backupset;
RMAN> recover datafile 6;
SQL> alter database open;

11. RMAN備份維護指令

• validate backupset

RMAN> validate backupset 5;

• restore...validate

RMAN> restore tablespace users validate;
RMAN> restore datafile '/data/oracle/oracent/system01.dbf' validate;

• restore preview

RMAN> restore database preview
RMAN> restore tablespace sysaux preview;
RMAN> restore datafile 5 preview;

• list

RMAN> list;
RMAN> list backupset;
RMAN> list backupset 5;
RMAN> list backup of tablespace users;
RMAN> list backup of datafile 1;
RMAN> list backup of archivelog all;
RMAN> list backup of archivelog from time ='sysdate-2';
RMAN> list backup of controlfile;
RMAN> list backup of spfile;
RMAN> list copy of controlfile;
RMAN> list backup summary;

• report

RMAN> report schema;
RMAN> report need backup;