ORACLE 11G通過SCN做增量備份修復standby庫詳細過程

背景描述：Oracle 的standby庫後臺alert報錯，如下：

ORA-00354: corrupt redo log block header
ORA-00353: log corruption near block 10240change 11125950022 time 05/08/2015 22:00:41
ORA-00334: archived log:'/data/oracle/oradgdata/standby_archive/1_32350_821708334.dbf'
Recovery interrupted!
Wed May 13 13:26:08 2015
Trace dumping is performingid=[cdmp_20150513132608]
Wed May 13 13:26:08 2015
Sweep [inc][273026]: completed
Recovered data files to a consistent stateat change 11125946527
Errors in file/oracle/app/oracle/diag/rdbms/pddgunq/powerdes/trace/powerdes_pr00_21813.trc:
ORA-00354: corrupt redo log block header
ORA-00353: log corruption near block 10240change 11125950022 time 05/08/2015 22:00:41
ORA-00334: archived log:'/data/oracle/oradgdata/standby_archive/1_32350_821708334.dbf'
MRP0: Background Media Recovery processshutdown (powerdes)
Wed May 13 13:39:58 2015
Standby controlfile consistent with primary
RFS[3]: Selected log 5 for thread 1sequence 32488 dbid -903205653 branch 821708334
Wed May 13 13:39:58 2015
Archived Log entry 24243 added for thread 1sequence 32487 ID 0xca2ab4eb dest 3:

分析：

這表明standby已經無法應用歸檔了，在32350這個歸檔日誌錯誤，但是歸檔日誌檔案/data/oracle/oradgdata/standby_archive/1_32350_821708334.dbf存在，只是無法應用了。

以前使用Duplicate target database命令恢復線上oracle datagard備庫，但是這個是整個庫恢復，消耗時間比較長；想到oracle還有一種可以基於scn的方式來恢復standby庫，所以才有那個基於scn的增量備份來恢復standby庫。

1，去備庫查詢下未歸檔的記錄的起始值

SQL> select min(sequence#) fromv$archived_log where applied='NO'

2 ;

MIN(SEQUENCE#)

--------------

32350

SQL> select max(sequence#) fromv$archived_log where applied='NO';

MAX(SEQUENCE#)

--------------

32508

SQL>

看到從32350到32508，都是沒有應用的歸檔日誌記錄。

----------------------------------------------------------------------------------------------------------------

原部落格地址： http://blog.itpub.net/26230597/viewspace-1661116/
原作者：黃杉 (mchdba)
----------------------------------------------------------------------------------------------------------------

2，去主庫查詢增量備份需要的SCN號

去主庫查詢下傳輸過來的32530歸檔日誌所對應的scn號，查詢sql如下：

SELECT SEQUENCE#,to_char(FIRST_CHANGE#)fc,to_char(NEXT_CHANGE#)nc FROM v$archived_log WHERE SEQUENCE# > 32349 ORDERBY 1;

SELECT SEQUENCE#,to_char(FIRST_CHANGE#)fc,to_char(NEXT_CHANGE#)nc FROM v$archived_log WHERE SEQUENCE# = 32350 ORDER BY1;

SQL> SELECT SEQUENCE#,to_char(FIRST_CHANGE#)fc,to_char(NEXT_CHANGE#)nc FROM v$archived_log WHERE SEQUENCE# = 32350 ORDER BY1;

SEQUENCE# FC NC

-------------------------------------------------- ----------------------------------------

32350 11125946510 11125975101

SQL>

我們看到主庫32350歸檔日誌在主庫對應的FIRST_CHANGE#的scn號是11125946510

去主備庫檢查下scn，可以看到彼此scn不一致：

select to_char(current_scn) scn fromv$database;

主庫：

SQL> select to_char(current_scn) scnfrom v$database;

SCN

----------------------------------------

11134239189

SQL>

備庫：

SQL> select to_char(current_scn) scnfrom v$database;

SCN

----------------------------------------

11125946526

SQL>

也可以按照備庫的最後一個scn號來在主庫上進行增量備份，不過為了保險起見，我們以主庫的scn號為準，那麼就需要基於11125946510這個scn號去主庫上做增量備份。

3，在主庫上設定log_archive_dest_state_2值，主庫的日誌無法歸檔到備庫.

先查詢自己的歸檔目錄select * from V$ARCHIVE_DEST;

PS：select * from V$ARCHIVE_DEST;查詢到STATUS為VALID的，然後DESTINATION為PD_DG（這裡是備庫的標識）的dest編號就是往備庫傳輸歸檔日誌的；另外一個有目錄/oracle/app/oracle/flash_recovery_area/archivelog的就是主庫本身的歸檔日誌dest，這個目錄儲存的就是主庫本身的歸檔日誌

ALTER system SET log_archive_dest_state_2 ='defer';

SQL> ALTER system SETlog_archive_dest_state_2 = 'defer';

System altered.

SQL>

4，去備庫先停止備庫應用

ALTER DATABASE RECOVER MANAGED STANDBYDATABASE CANCEL;

SQL> ALTER DATABASE RECOVER MANAGEDSTANDBY DATABASE CANCEL;

ALTER DATABASE RECOVER MANAGED STANDBYDATABASE CANCEL

*

ERROR at line 1:

ORA-16136: Managed Standby Recovery notactive

SQL>

看到standby已經是not active了，所以不需要執行了。

5，在主庫執行備份

基於SCN增量備份：

backup device type disk incremental fromscn 11125946510 database format '/home/oracle/db_incre%U.bbk';

RMAN> backup device type diskincremental from scn 11125946510 database format'/home/oracle/db_incre%U.bbk';

Starting backup at 13-MAY-15

using target database control file insteadof recovery catalog

allocated channel: ORA_DISK_1

channel ORA_DISK_1: SID=212 devicetype=DISK

backup will be obsolete on date20-MAY-15

archived logs will not be kept or backed up

channel ORA_DISK_1: starting full datafilebackup set

channel ORA_DISK_1: specifying datafile(s) inbackup set

input datafile file number=00005name=/home/oradata/powerdes/powerdesk01.dbf

input datafile file number=00003name=/home/oradata/powerdes/undotbs01.dbf

input datafile file number=00006name=/home/oradata/powerdes/plas01.dbf

input datafile file number=00001name=/home/oradata/powerdes/system01.dbf

input datafile file number=00002name=/home/oradata/powerdes/sysaux01.dbf

input datafile file number=00007name=/home/oradata/powerdes/pl01.dbf

input datafile file number=00011 name=/home/oradata/powerdes/plcrm01.dbf

input datafile file number=00004name=/home/oradata/powerdes/users01.dbf

input datafile file number=00008name=/home/oradata/powerdes/help01.dbf

input datafile file number=00009name=/home/oradata/powerdes/adobelc01.dbf

input datafile file number=00010name=/home/oradata/powerdes/sms01.dbf

channel ORA_DISK_1: starting piece 1 at13-MAY-15

channel ORA_DISK_1: finished piece 1 at13-MAY-15

piecehandle=/home/oracle/db_increi3q6s13g_1_1.bbk tag=TAG20150513T202207comment=NONE

channel ORA_DISK_1: backup set complete,elapsed time: 00:09:45

using channel ORA_DISK_1

backup will be obsolete on date 20-MAY-15

archived logs will not be kept or backed up

channel ORA_DISK_1: starting full datafilebackup set

channel ORA_DISK_1: specifying datafile(s)in backup set

including current control file in backupset

channel ORA_DISK_1: starting piece 1 at13-MAY-15

channel ORA_DISK_1: finished piece 1 at13-MAY-15

piecehandle=/home/oracle/db_increi4q6s1lp_1_1.bbk tag=TAG20150513T202207comment=NONE

channel ORA_DISK_1: backup set complete,elapsed time: 00:00:01

Finished backup at 13-MAY-15

RMAN>

檢視是否備份成功

RMAN> list backup of database;

......

BS Key Type LV Size Device TypeElapsed Time Completion Time

------- ---- -- ---------- ----------------------- ---------------

3343 Incr 2.41G DISK 00:09:35 13-MAY-15

BP Key: 3343 Status:AVAILABLE Compressed: NO Tag: TAG20150513T202207

Piece Name: /home/oracle/db_increi3q6s13g_1_1.bbk

Keep: NOLOGS Until:20-MAY-15

List of Datafiles in backup set 3343

File LV Type Ckp SCN CkpTime Name

---- -- ---- ---------- --------- ----

1 Incr 11134253054 13-MAY-15/home/oradata/powerdes/system01.dbf

2 Incr 11134253054 13-MAY-15/home/oradata/powerdes/sysaux01.dbf

3 Incr 11134253054 13-MAY-15/home/oradata/powerdes/undotbs01.dbf

4 Incr 11134253054 13-MAY-15/home/oradata/powerdes/users01.dbf

5 Incr 11134253054 13-MAY-15/home/oradata/powerdes/powerdesk01.dbf

6 Incr 11134253054 13-MAY-15/home/oradata/powerdes/plas01.dbf

7 Incr 11134253054 13-MAY-15/home/oradata/powerdes/pl01.dbf

8 Incr 11134253054 13-MAY-15/home/oradata/powerdes/help01.dbf

9 Incr 11134253054 13-MAY-15/home/oradata/powerdes/adobelc01.dbf

10 Incr 11134253054 13-MAY-15/home/oradata/powerdes/sms01.dbf

11 Incr 11134253054 13-MAY-15/home/oradata/powerdes/plcrm01.dbf

RMAN>

去檢視備份的檔案，看到2個bbk，表示備份成功，如下所示：

[oracle@localhost ~]$ ll db*

-rw-r----- 1 oracle oinstall 2591825920 May13 20:31 db_increi3q6s13g_1_1.bbk

-rw-r----- 1 oracle oinstall 19234816 May 13 20:31db_increi4q6s1lp_1_1.bbk

[oracle@localhost ~]$

6，在主庫重新生成控制檔案

ALTER DATABASE CREATE standby controlfileAS '/home/oracle/standby.ctl';

SQL> ALTER DATABASE CREATE standbycontrolfile AS '/home/oracle/standby.ctl';

Database altered.

SQL>

7，copy備份檔案到備庫

[oracle@localhost ~]$ scp db*/home/oracle/standby.ctl root@192.168.xx.xx:/data/oracle/backup/restore/

root@192.168.xx.xx's password:

db_increi3q6s13g_1_1.bbk 100% 2472MB 95.1MB/s 00:26

db_increi4q6s1lp_1_1.bbk 100% 18MB 18.3MB/s 00:00

standby.ctl 100% 18MB 18.3MB/s 00:00

[oracle@localhost ~]$

8，在備庫操作，準備好rman前環境

關閉備庫

SQL> shutdown immediate

ORA-01109: database not open

Database dismounted.

ORACLE instance shut down.

SQL>

然後啟動到nomont狀態

SQL> startup nomount

ORACLE instance started.

Total System Global Area 5344731136 bytes

Fixed Size 2213136 bytes

Variable Size 3489663728 bytes

Database Buffers 1811939328 bytes

Redo Buffers 40914944 bytes

SQL>

9，備庫上通過rman恢復控制檔案

restore controlfile from'/data/oracle/backup/restore/standby.ctl';

RMAN> restore controlfile from'/data/oracle/backup/restore/standby.ctl';

Starting restore at 13-MAY-15

using channel ORA_DISK_1

RMAN-00571:===========================================================

RMAN-00569: =============== ERROR MESSAGESTACK FOLLOWS ===============

RMAN-00571:===========================================================

RMAN-03002: failure of restore command at05/13/2015 20:44:02

RMAN-06172: no AUTOBACKUP found orspecified handle is not a valid copy or piece

RMAN>

是目錄許可權的問題，需要設定成oracle使用者

[root@localhost restore]# ll

total 2568604

-rw-r----- 1 root root 2591825920 May 1320:35 db_increi3q6s13g_1_1.bbk

-rw-r----- 1 root root 19234816 May 13 20:35db_increi4q6s1lp_1_1.bbk

-rw-r----- 1 root root 19185664 May 13 20:35 standby.ctl

[root@localhost restore]# chown -Roracle.dba *

[root@localhost restore]#

RMAN> restore controlfile from'/data/oracle/backup/restore/standby.ctl';

Starting restore at 13-MAY-15

using target database control file insteadof recovery catalog

allocated channel: ORA_DISK_1

channel ORA_DISK_1: SID=386 devicetype=DISK

channel ORA_DISK_1: copied control filecopy

output filename=/home/oradata/powerdes/control01.ctl

output filename=/oracle/app/oracle/flash_recovery_area/powerdes/control02.ctl

Finished restore at 13-MAY-15

RMAN>

10，備庫上rman執行catalog操作

先將資料庫載入到mount，然後再執行catalog start with '/data/oracle/backup/restore'恢復

RMAN> alter database mount

2> ;

using target database control file insteadof recovery catalog

database mounted

RMAN> catalog start with '/data/oracle/backup/restore'

RMAN> catalog start with'/data/oracle/backup/restore';

Starting implicit crosscheck backup at13-MAY-15

allocated channel: ORA_DISK_1

channel ORA_DISK_1: SID=867 devicetype=DISK

Crosschecked 51 objects

Finished implicit crosscheck backup at13-MAY-15

Starting implicit crosscheck copy at13-MAY-15

using channel ORA_DISK_1

Finished implicit crosscheck copy at13-MAY-15

searching for all files in the recoveryarea

cataloging files...

no files cataloged

searching for all files that match thepattern /data/oracle/backup/restore

List of Files Unknown to the Database

=====================================

File Name:/data/oracle/backup/restore/db_increi3q6s13g_1_1.bbk

File Name:/data/oracle/backup/restore/db_increi4q6s1lp_1_1.bbk

File Name: /data/oracle/backup/restore/standby.ctl

Do you really want to catalog the abovefiles (enter YES or NO)? yes

cataloging files...

cataloging done

List of Cataloged Files

=======================

File Name: /data/oracle/backup/restore/db_increi3q6s13g_1_1.bbk

File Name:/data/oracle/backup/restore/db_increi4q6s1lp_1_1.bbk

File Name:/data/oracle/backup/restore/standby.ctl

RMAN>

11，備庫上rman執行restore操作

RECOVER DATABASE NOREDO; PS：加NOREDO不用線上日誌

RMAN> RECOVER DATABASE NOREDO;

Starting recover at 13-MAY-15

using channel ORA_DISK_1

channel ORA_DISK_1: starting incrementaldatafile backup set restore

channel ORA_DISK_1: specifying datafile(s)to restore from backup set

destination for restore of datafile 00001:/home/oradata/powerdes/system01.dbf

destination for restore of datafile 00002:/home/oradata/powerdes/sysaux01.dbf

destination for restore of datafile 00003:/home/oradata/powerdes/undotbs01.dbf

destination for restore of datafile 00004:/home/oradata/powerdes/users01.dbf

destination for restore of datafile 00005:/home/oradata/powerdes/powerdesk01.dbf

destination for restore of datafile 00006:/home/oradata/powerdes/plas01.dbf

destination for restore of datafile 00007:/home/oradata/powerdes/pl01.dbf

destination for restore of datafile 00008:/home/oradata/powerdes/help01.dbf

destination for restore of datafile 00009:/home/oradata/powerdes/adobelc01.dbf

destination for restore of datafile 00010:/home/oradata/powerdes/sms01.dbf

destination for restore of datafile 00011:/home/oradata/powerdes/plcrm01.dbf

channel ORA_DISK_1: reading from backuppiece /data/oracle/backup/restore/db_increi3q6s13g_1_1.bbk

channel ORA_DISK_1: piecehandle=/data/oracle/backup/restore/db_increi3q6s13g_1_1.bbktag=TAG20150513T202207

channel ORA_DISK_1: restored backup piece 1

channel ORA_DISK_1: restore complete,elapsed time: 00:05:45

Finished recover at 13-MAY-15

RMAN>

後臺alert日誌輸出為：

Using STANDBY_ARCHIVE_DEST parameterdefault value as /data/oracle/oradgdata/standby_archive

Wed May 13 20:54:42 2015

Incremental restore complete of datafile 4/home/oradata/powerdes/users01.dbf

checkpoint is 11134253054

last deallocation scn is 10905979642

Wed May 13 20:54:53 2015

Incremental restore complete of datafile 8/home/oradata/powerdes/help01.dbf

checkpoint is 11134253054

last deallocation scn is 9881798870

Wed May 13 20:55:08 2015

Incremental restore complete of datafile 9/home/oradata/powerdes/adobelc01.dbf

checkpoint is 11134253054

Wed May 13 20:55:21 2015

Incremental restore complete of datafile 10/home/oradata/powerdes/sms01.dbf

checkpoint is 11134253054

Wed May 13 20:55:34 2015

Incremental restore complete of datafile 11/home/oradata/powerdes/plcrm01.dbf

checkpoint is 11134253054

Incremental restore complete of datafile 7/home/oradata/powerdes/pl01.dbf

checkpoint is 11134253054

last deallocation scn is 10905470965

Wed May 13 20:56:24 2015

db_recovery_file_dest_size of 15360 MB is0.00% used. This is a

user-specified limit on the amount of spacethat will be used by this

database for recovery-related files, anddoes not reflect the amount of

space available in the underlyingfilesystem or ASM diskgroup.

Wed May 13 20:56:26 2015

Incremental restore complete of datafile 2/home/oradata/powerdes/sysaux01.dbf

checkpoint is 11134253054

last deallocation scn is 10906515871

Wed May 13 20:56:43 2015

Incremental restore complete of datafile 1/home/oradata/powerdes/system01.dbf

checkpoint is 11134253054

last deallocation scn is 10825889348

Incremental restore complete of datafile 6/home/oradata/powerdes/plas01.dbf

checkpoint is 11134253054

last deallocation scn is 10905977986

Incremental restore complete of datafile 3/home/oradata/powerdes/undotbs01.dbf

checkpoint is 11134253054

last deallocation scn is 10908635608

Wed May 13 20:59:58 2015

Incremental restore complete of datafile 5/home/oradata/powerdes/powerdesk01.dbf

checkpoint is 11134253054

last deallocation scn is 10906663694

12，備庫上重新啟動應用日誌

ALTER DATABASE recover managed standby DATABASE disconnect FROM SESSION;

SQL>ALTER DATABASE recover managed standby DATABASE disconnect FROM SESSION;

Databasealtered.

SQL>

後臺alert日誌輸出為：

ALTER DATABASE recover managed standby DATABASE disconnect FROM SESSION

Attemptto start background Managed Standby Recovery process (powerdes)

WedMay 13 21:04:41 2015

MRP0started with pid=19, OS id=23729

MRP0:Background Managed Standby Recovery process started (powerdes)

started logmerger process

WedMay 13 21:04:46 2015

ManagedStandby Recovery not using Real Time Apply

ParallelMedia Recovery started with 16 slaves

Waitingfor all non-current ORLs to be archived...

Allnon-current ORLs have been archived.

MediaRecovery Waiting for thread 1 sequence 32510

Completed:ALTER DATABASE recover managed standby DATABASE disconnect FROM SESSION

13，主庫操作往備庫傳輸歸檔日誌

命令為：

ALTER system SET log_archive_dest_state_2 ='enable';

去備庫後臺alert日誌資訊為：

Standby controlfile consistent with primary

SRL log 4 needs clearing because log hasnot been created

Errors in file/oracle/app/oracle/diag/rdbms/pddgunq/powerdes/trace/powerdes_rfs_23767.trc: