nbu怎样恢复oracle数据库

2、安装数据库系统

3、安装nbu软件及配置

--安装客户端

上传SYMCnbclient_Linux-RedHat2.6.18_7.5.0.2.tar编辑本机hosts

vi /etc/hosts

添加：192.168.99.252 hdnbu1

解压：

[root@scmtest u01]# tar -xvf SYMCnbclient_Linux-RedHat2.6.18_7.5.0.2.tar[root@scmtest u01]# cd SYMCnbclient_Linux-RedHat2.6.18_7.5.0.2[root@scmtest SYMCnbclient_Linux-RedHat2.6.18_7.5.0.2]# ./install过程当中要输入备份服务端：hdnbu1

然后再用oracle用户执行：

[root@scmtest bin]# su - oracle

[oracle@scmtest ~]$ cd /usr/openv/netbackup/bin[oracle@scmtest bin]$ ./oracle_link

Thu Feb 21 14:03:02 CST 2013

All Oracle instances should be shutdown before running this script.

Please log into the Unix system as the Oracle owner for running this scriptDo you want to continue? (y/n) [n] y

LIBOBK path: /usr/openv/netbackup/bin

ORACLE_HOME: /software/oracle/product/10.2.0/db1Oracle version: 10.2.0.5.0

Platform type: x86_64

Linking LIBOBK:

ln -s /usr/openv/netbackup/bin/libobk.so64 /software/oracle/product/10.2.0/db1/lib/libobk.soDone

Please check the trace file located in /tmp/make_trace.23602to make sure the linking process was successful.

4、编辑初始参数文件

这时候，需要进行修改，比如原来用的什么盘，现在用什么盘等编辑后如下：

*._addm_auto_enable=FALSE

*._b_tree_bitmap_plans=FALSE

*._db_block_numa=1

*._enable_NUMA_optimization=FALSE

*._optimizer_cartesian_enabled=FALSE

*._optimizer_skip_scan_enabled=FALSE

*.archive_lag_target=1800

*.audit_file_dest='/software/oracle/admin/ncerp/adump'

*.background_dump_dest='/software/oracle/admin/ncerp/bdump'

*.compatible='10.2.0.5.0'

*.control_files='/u01/ncerp/control/control01.ctl','/u01/ncerp/control/control02.ctl'

*.core_dump_dest='/software/oracle/admin/ncerp/cdump'

*.db_block_size=8192

*.db_cache_size=4294967296

*.db_create_file_dest='/u01/ncerp/flash_recover'

*.db_domain=''

*.db_file_multiblock_read_count=16

*.db_name='ncerp'

*.java_pool_size=33554432

*.job_queue_processes=10

*.large_pool_size=33554432

*.log_archive_dest_1='location=/u01/ncerp/arch'

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

*.max_dump_file_size='1024'

*.open_cursors=1000

*.optimizer_dynamic_sampling=4

*.optimizer_index_cost_adj=40

*.parallel_max_servers=5

*.pga_aggregate_target=2147483648

*.processes=800

*.recyclebin='OFF'

*.shared_pool_size=838860800

*.undo_management='AUTO'

*.undo_tablespace='UNDOTBS1'

5、根据参数文件，建相应的目录

[oracle@scmtest ~]$ mkdir -p /u01/ncerp/control/[oracle@scmtest u01]$ mkdir -p /app/oracle/admin/ncerp/cdump[oracle@scmtest u01]$ mkdir -p /app/oracle/admin/ncerp/bdump[oracle@scmtest u01]$ mkdir -p /app/oracle/admin/ncerp/adump[oracle@scmtest u01]$ mkdir -p /app/oracle/admin/ncerp/pfile[oracle@scmtest u01]$ mkdir -p /u01/ncerp/flash_recover[oracle@scmtest u01]$ mkdir -p /u01/ncerp/arch[oracle@scmtest ncerp]$ mkdir oradata

注意：这些要与参数文件对应，要不然无法启动！

6、恢复控制文件

export ORACLE_SID=ncerp

[oracle@scmtest u01]$ rman target /

Recovery Manager: Release 10.2.0.5.0 - Production on Thu Feb 21 14:17:10 2013Copyright (c) 1982, 2007, Oracle. All rights reserved.

connected to target database (not started)RMAN>startup nomount pfile='/u01/ncpfile.ora'Oracle instance started

Total System Global Area5217714176 bytesFixed Size 2103536 bytesVariable Size905971472 bytesDatabase Buffers4294967296 bytesRedo Buffers 14671872 bytes从nbu的服务器上查询要恢复的控制文件

hdnbu1:/tmp # bplist -C RD162 -t 4 -R -b -l / | more-rw-rw---- oracledba 10747904 Mar 28 11:26 /c-3383507379-20130328-0b-rw-rw---- oracledba 11010048 Mar 28 11:25 /c-2177845250-20130328-0d-rw-rw---- oracledba 10747904 Mar 28 11:25 /cntrl_9202_1_811250475-rw-rw---- oracledba 10747904 Mar 28 11:25 /cntrl_7788_1_811250456-rw-rw---- oracledba 10747904 Mar 28 11:25 /c-3383507379-20130328-0a-rw-rw---- oracledba 11010048 Mar 28 11:24 /c-2177845250-20130328-0c-rw-rw---- oracledba 114032640 Mar 28 11:17 /al_7786_1_811249961-rw-rw---- oracledba 123207680 Mar 28 11:17 /al_7785_1_811249961-rw-rw---- oracledba 3670016 Mar 28 11:17 /al_9200_1_811249960-rw-rw---- oracledba 3932160 Mar 28 11:17 /al_9199_1_811249959-rw-rw---- oracledba 11010048 Mar 28 09:25 /c-2177845250-20130328-0b-rw-rw---- oracledba 10747904 Mar 28 09:25 /c-3383507379-20130328-09-rw-rw---- oracledba 10747904 Mar 28 09:25 /cntrl_9197_1_811243259-rw-rw---- oracledba 10747904 Mar 28 09:25 /cntrl_7783_1_811243253RMAN>run

2>{

3>allocate channel c1 type 'sbt_tape'

4>send 'NB_ORA_CLIENT=RD162'

5>restore controlfile from '/cntrl_24145_1_828414748'6>release channel c1

7>}

using target database control file instead of recovery catalogallocated channel: c1

channel c1: sid=874 devtype=SBT_TAPE

channel c1: Veritas NetBackup for Oracle - Release 7.5 (2012050902)sent command to channel: c1

Starting restore at 2013-03-28 12:41:30

channel c1: restoring control file

channel c1: restore complete, elapsed time: 00:05:16output filename=/u02/ncerp/control/control01.ctloutput filename=/u02/ncerp/control/control02.ctlFinished restore at 2013-03-28 12:46:47

released channel: c1

7、还原数据库文件

在这里要查一下数据库是否已经mount

SQL>select status from v$instance

STATUS

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

MOUNTED

RMAN>run {

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE'

send 'NB_ORA_CLIENT=RD162'

set newname for datafile '/backup/ncerp/oradata/system01.dbf' to '/u02/ncerp/oradata/system01.dbf' set newname for datafile '/backup/ncerp/oradata/system02.dbf' to '/u02/ncerp/oradata/system02.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_index03_1.dbf' to '/u02/ncerp/oradata/nnc_index03_1.dbf'set newname for datafile '/backup/ncerp/oradata/undotbs1.dbf' to '/u02/ncerp/oradata/undotbs1.dbf' set newname for datafile '/backup/ncerp/oradata/sysaux01.dbf' to '/u02/ncerp/oradata/sysaux01.dbf' set newname for datafile '/backup/ncerp/oradata/users01.dbf' to '/u02/ncerp/oradata/users01.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data01_1.dbf' to '/u02/ncerp/oradata/nnc_data01_1.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data01_2.dbf' to '/u02/ncerp/oradata/nnc_data01_2.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data01_3.dbf' to '/u02/ncerp/oradata/nnc_data01_3.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data01_4.dbf' to '/u02/ncerp/oradata/nnc_data01_4.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data02_1.dbf' to '/u02/ncerp/oradata/nnc_data02_1.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data02_2.dbf' to '/u02/ncerp/oradata/nnc_data02_2.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data03_1.dbf' to '/u02/ncerp/oradata/nnc_data03_1.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data03_2.dbf' to '/u02/ncerp/oradata/nnc_data03_2.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data03_3.dbf' to '/u02/ncerp/oradata/nnc_data03_3.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data03_4.dbf' to '/u02/ncerp/oradata/nnc_data03_4.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_index02_1.dbf' to '/u02/ncerp/oradata/nnc_index02_1.dbf'set newname for datafile '/backup/ncerp/oradata/nnc_index01_1.dbf' to '/u02/ncerp/oradata/nnc_index01_1.dbf'set newname for datafile '/backup/ncerp/oradata/nnc_index01_2.dbf' to '/u02/ncerp/oradata/nnc_index01_2.dbf'set newname for datafile '/backup/ncerp/oradata/nnc_index01_3.dbf' to '/u02/ncerp/oradata/nnc_index01_3.dbf'set newname for datafile '/backup/ncerp/oradata/nnc_index01_4.dbf' to '/u02/ncerp/oradata/nnc_index01_4.dbf'set newname for datafile '/backup/ncerp/oradata/iufo01.dbf'to '/u02/ncerp/oradata/iufo01.dbf' set newname for datafile '/backup/ncerp/oradata/iufo02.dbf'to '/u02/ncerp/oradata/iufo02.dbf' set newname for datafile '/backup/ncerp/oradata/nnc_data03_5.dbf' to '/u02/ncerp/oradata/nnc_data03_5.dbf' Restore database

Switch datafile all

}

vi res_test.sh 内容如下：

export ORACLE_SID=ncerp

rman target / <run {

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE'

send 'NB_ORA_CLIENT=RD162'

做一次恢复测试看看

说明需要3556这个归档日志

所以恢复时，从这个日志进行恢复！

也可以采用后台运行方式

9、数据库恢复

SQL>

cancel

说明，如果上面日志可能有点大的时候，我们需要调整，比如说删除日志组等，然后还重新添加日志到新位置10、打开数据库

SQL>alter database open resetlogs

Database altered.

11、创建临时表空间

2.用pfile启动到MOUNT，恢复控制文件

3.根据控制文件恢复数据文件

4.resetlog打开数据库

大致流程就这样了！