RMAN数据库恢复失败一则

问题： 这是一个从RAC环境的数据库的RAMN备份恢复到一个单机数据库的操作。 当恢复数据文件和恢复正常，但在open数据库时出报下面的错误。 --rman备份恢复操作 #创建参数文件 cd $ORACLE_HOME/dbs $cat initntracdb.ora *.archive_lag_target=0 *.compatible

问题：

这是一个从RAC环境的数据库的RAMN备份恢复到一个单机数据库的操作。

当恢复数据文件和恢复正常，但在open数据库时出报下面的错误。

--rman备份恢复操作

#创建参数文件

cd $ORACLE_HOME/dbs

$cat initntracdb.ora

*.archive_lag_target=0

*.compatible='11.2.0.4.0'

*.control_files='/u01/oracle/oradata/ntracdb/controlfile1.dbf','/u01/oracle/oradata/ntracdb/controlfile2.dbf'

*.db_block_size=8192

*.db_create_file_dest='/u01/oracle/oradata/ntracdb'

*.db_name='ntracdb'

*.db_recovery_file_dest='/u01/oracle/fast_recovery_area'

*.db_recovery_file_dest_size=299000M

*.db_unique_name='ntracdb'

*.dg_broker_start=true

*.local_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=nticket3)(PORT=1521))'

*.log_archive_format='%t_%s_%r.dbf'

*.log_archive_max_processes=4

*.log_archive_min_succeed_dest=1

*.log_archive_trace=0

*.log_file_name_convert='null','null'

*.nls_language='SIMPLIFIEDCHINESE'

*.nls_territory='CHINA'

*.open_cursors=300

*.pga_aggregate_target=429496729

*.processes=600

*.remote_login_passwordfile='exclusive'

*.sga_max_size=3435973836

*.sga_target=3221225472

*.standby_file_management='auto'

*.undo_tablespace='UNDOTBS1'

rman target /

startup nomount;

restore controlfile from'/home/oracle/rmanbak/ncnnf0_TAG20141110T011010_0.1205.863228449'; --首先恢复控制文件

alter database mount;

catalog start with'/home/oracle/rmanbak/'; --批量登记拷过来的rman备份，假设拷过来的备份放到了/u01/rmanbak/目录

list backup; --查看要恢复的是不是这个备份文件

run {

set newname for datafile'+DATA01/ntracdb/datafile/users.295.855410331' to'/u01/oracle/oradata/ntracdb/users.295.855410331';

set newname for datafile'+DATA01/ntracdb/datafile/undotbs1.263.855410331' to'/u01/oracle/oradata/ntracdb/undotbs1.263.855410331';

set newname for datafile'+DATA01/ntracdb/datafile/sysaux.264.855410331' to'/u01/oracle/oradata/ntracdb/sysaux.264.855410331';

set newname for datafile'+DATA01/ntracdb/datafile/system.265.855410331' to'/u01/oracle/oradata/ntracdb/system.265.855410331';

set newname for datafile'+DATA01/ntracdb/datafile/undotbs2.293.855410453' to'/u01/oracle/oradata/ntracdb/undotbs2.293.855410453';

set newname for datafile'+DATA01/ntracdb/datafile/undotbs3.292.855410453' to'/u01/oracle/oradata/ntracdb/undotbs3.292.855410453';

set newname for datafile'+DATA01/ntracdb/datafile/sysaux.257.857772301' to '/u01/oracle/oradata/ntracdb/sysaux.257.857772301';

set newname for datafile'+DATA01/ntracdb/datafile/strategy.256.858008275' to'/u01/oracle/oradata/ntracdb/strategy.256.858008275'

restore database;

switch datafile all;

recover database;

}

--打开数据库时报错

$sqlplus / as sysdba

SQL> alter database open;

alter database open

*

第 1 行出现错误

ORA-03113：通信通道的文件结尾

进程 ID ：6988

回话 ID：191 序列号：3

--查看日志

Thu Nov 13 10:13:20 2014

alter database open

Data Guard Brokerinitializing...

Data Guard Brokerinitialization complete

Data Guard: verifying databaseprimary role...

Thu Nov 13 10:13:20 2014

LGWR: STARTING ARCH PROCESSES

Thu Nov 13 10:13:20 2014

ARC0 started with pid=21, OSid=26949

ARC0: Archival started

LGWR: STARTING ARCH PROCESSESCOMPLETE

ARC0: STARTING ARCH PROCESSES

LGWR: Primary database is inMAXIMUM AVAILABILITY mode

LGWR: DestinationLOG_ARCHIVE_DEST_1 is not serviced by LGWR

LGWR: Minimum of 1 LGWR standbydatabase required

Errors in file/u01/oracle/diag/rdbms/ntracdb/ntracdb/trace/ntracdb_lgwr_26870.trc:

ORA-16072: a minimum of onestandby database destination is required

Thu Nov 13 10:13:21 2014

ARC1 started with pid=22, OSid=26953

LGWR (ospid: 26870):terminating the instance due to error 16072

Thu Nov 13 10:13:21 2014

System statedump requested by (instance=1, osid=26870 (LGWR)), summary=[abnormal instancetermination].

System Statedumped to trace file/u01/oracle/diag/rdbms/ntracdb/ntracdb/trace/ntracdb_diag_26846_20141113101321.trc

Dumpingdiagnostic data in directory=[cdmp_20141113101321], requested by (instance=1,osid=26870 (LGWR)), summary=[abnormal instance termination].

Instanceterminated by LGWR, pid = 26870

原因：

可能是控制文件备份时失败所致

解决办法：

重建控制文件，然后再打开数据库

STARTUP NOMOUNT

CREATE CONTROLFILE REUSEDATABASE "ntracdb" NORESETLOGS FORCE LOGGING ARCHIVELOG

MAXLOGFILES 16

MAXLOGMEMBERS 3

MAXDATAFILES 100

MAXINSTANCES 8

MAXLOGHISTORY 9344

LOGFILE

GROUP 1'/u01/oracle/oradata/ntracdb/NTRACDB/onlinelog/o1_mf_1_b682j5nk_.log' SIZE 200M,

GROUP 2 '/u01/oracle/oradata/ntracdb/NTRACDB/onlinelog/o1_mf_2_b682j7gw_.log' SIZE 200M,

GROUP 3'/u01/oracle/oradata/ntracdb/NTRACDB/onlinelog/o1_mf_3_b682j98k_.log' SIZE 200M,

GROUP 4'/u01/oracle/oradata/ntracdb/NTRACDB/onlinelog/o1_mf_4_b682jc2t_.log' SIZE 200M

-- STANDBY LOGFILE

DATAFILE

'/u01/oracle/oradata/ntracdb/users.295.855410331',

'/u01/oracle/oradata/ntracdb/undotbs1.263.855410331',

'/u01/oracle/oradata/ntracdb/sysaux.264.855410331',

'/u01/oracle/oradata/ntracdb/system.265.855410331',

'/u01/oracle/oradata/ntracdb/undotbs2.293.855410453',

'/u01/oracle/oradata/ntracdb/undotbs3.292.855410453',

'/u01/oracle/oradata/ntracdb/sysaux.257.857772301',

'/u01/oracle/oradata/ntracdb/strategy.256.858008275',

'/u01/oracle/oradata/ntracdb/strategy.302.858008423'

CHARACTER SET ZHS16GBK;

SQL> RECOVER DATABASE;

ORA-00283: 恢复会话因错误而取消

ORA-00264: 不要求恢复

--此时可以正常打开数据库

SQL> alter database open;

数据库已更改。

#创建临时表空间

CREATE TEMPORARY TABLESPACE temp TEMPFILE'/u01/oracle/oradata/ntracdb/temp01.dbf'

SIZE 20M REUSE

EXTENT MANAGEMENT LOCAL UNIFORM SIZE 16M;