恢复例子

 

1.基于scn做不完全恢复
Select Recid,
       Stamp,
       Thread#,
       Sequence#,
       First_Change# First_Time,
       Next_Change#
  From V$archived_Log
 Where Resetlogs_Change# = (Select Resetlogs_Change#
                              From V$database_Incarnation
                             Where Status = 'CURRENT');
RECID STAMP     THREAD# SEQUENCE#   FIRST_TIME NEXT_CHANGE#
60   788191628   1        1       1154977     1155349
61   788191760   1        2       1155349     1155413
62   788191840   1        3       1155413     1155453

 

这里我们只恢复到日志2,即找到日志2里最大的scn为1155413

run{
  set until scn 1155413;
  allocate channel c1 device type disk;
  restore database;
  recover database;
  release channel c1;
}

 

2.基于sequence做不完全恢复
run{
  set until sequence 3 thread 1;
  allocate channel c1 device type disk;
  restore database;
  recover database;
  release channel c1;
}

这里用到日志3前的日志(不包含日志3)

 

3.基于时间点的不完全恢复
Select Recid,
       Stamp,
       Thread#,
       Sequence#,
       First_Change# First_Time,
       Next_Change#,
       t.completion_time -- 这里可以查到某个日志归档完成的时间点,可以参考这个时间点恢复
  From V$archived_Log t
 Where Resetlogs_Change# = (Select Resetlogs_Change#
                              From V$database_Incarnation
                             Where Status = 'CURRENT');

run{
  set until time "to_date('2012-7-9 15:07:20','YYYY-MM-DD HH24:MI:SS')";
  allocate channel c1 device type disk;
  restore database;
  recover database;
  release channel c1;
}

 

4.表空间完全恢复

run{
  allocate channel c1 device type disk;
  restore tablespace tps_hxl01;
  recover tablespace tps_hxl01;
  release channel c1;
}