磁碟損壞，system及部分資料檔案丟失

Oracle 9.2.0.7     RAC    2 NODES . 

將rac production資料庫關閉， 修復損壞磁碟區， 將備份的datafile 檔案(這裡我們使用standby datafiles) restore 到修復後的磁碟中，然後將rac資料庫中一個例項開啟為mount 狀態 (當然前提是controlfile是沒有損壞的)， 在這個節點上做恢復，  recover  automatic  database  ,   介質恢復完成後， alter database open ;  開啟例項1的庫。 然後直接在節點2上 startup 開啟庫， 開啟過程需要較長時間 。 

SQL> startup mount;
ORACLE instance started.

Total System Global Area 1679366908 bytes
Fixed Size                   453372 bytes
Variable Size             620756992 bytes
Database Buffers         1056964608 bytes
Redo Buffers                1191936 bytes
Database mounted.

SQL> recover automatic database ;
Media recovery complete.

SQL>

 

節點1上在恢復的時候的log :   

DELL-RAC01$tail -f alert_orcl1.log
This instance was first to mount
LCK0 started with pid=19
Sun May  1 14:25:49 2011
Successful mount of redo thread 1, with mount id 1277774953
Sun May  1 14:25:49 2011
Database mounted in Shared Mode (CLUSTER_DATABASE=TRUE).
Completed: ALTER DATABASE   MOUNT
Sun May  1 14:25:57 2011
ALTER DATABASE RECOVER  automatic database  
Media Recovery Start
Sun May  1 14:29:22 2011
Recovery of Online Redo Log: Thread 2 Group 7 Seq 16023 Reading mem 0
  Mem# 0 errs 0: /ocfs_ctrl_redo/orcl/redo07.log
  Mem# 1 errs 0: /ocfs_data/orcl/redo07b.log
Sun May  1 14:29:22 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 38430 Reading mem 0
  Mem# 0 errs 0: /ocfs_ctrl_redo/orcl/redo01.log
  Mem# 1 errs 0: /ocfs_data/orcl/redo01b.log
Sun May  1 14:29:23 2011
Recovery of Online Redo Log: Thread 1 Group 4 Seq 38431 Reading mem 0
  Mem# 0 errs 0: /ocfs_ctrl_redo/orcl/redo04.log
  Mem# 1 errs 0: /ocfs_data/orcl/redo04b.log
Media Recovery Complete
Completed: ALTER DATABASE RECOVER  automatic database  

 

startup 開啟節點2上的庫時的 alert log ：  

DELL-RAC02$tail -f alert_orcl2.log
Reconfiguration complete
LCK0 started with pid=19
Sun May  1 14:32:04 2011
Successful mount of redo thread 2, with mount id 1277774953
Sun May  1 14:32:04 2011
Database mounted in Shared Mode (CLUSTER_DATABASE=TRUE).
Completed: ALTER DATABASE   MOUNT
Sun May  1 14:32:05 2011
ALTER DATABASE OPEN
Picked Lamport scheme to generate SCNs
Sun May  1 14:38:02 2011
LGWR: Primary database is in CLUSTER CONSISTENT mode
Sun May  1 14:38:17 2011
Thread 2 opened at log sequence 16023
  Current log# 7 seq# 16023 mem# 0: /ocfs_ctrl_redo/orcl/redo07.log
  Current log# 7 seq# 16023 mem# 1: /ocfs_data/orcl/redo07b.log
Successful open of redo thread 2
Sun May  1 14:38:17 2011
SMON: enabling cache recovery
Sun May  1 14:38:19 2011
Successfully onlined Undo Tablespace 10.
Sun May  1 14:38:19 2011
SMON: enabling tx recovery
Sun May  1 14:38:19 2011
Database Characterset is AL32UTF8
replication_dependency_tracking turned off (no async multimaster replication found)
Completed: ALTER DATABASE OPEN