进入 rman 工具

rman target /

查看 rman 配置

rman> show all;

修改rman 配置

数据库全备

rman> run {

allocate channel c1 type disk;

allocate channel c2 type disk;

backup incremental level 0 database format '/home/oracle/backup/full_%d_%s_%t.bak';

sql 'alter system archive log current';

crosscheck archivelog all;

backup archivelog all format '/home/oracle/backup/archlog_%d_%s_%t.bak' delete input;

backup spfile format '/home/oracle/backup/spfile_%d_%s_%t.bak';

backup current controlfile format '/home/oracle/backup/controlfile_%d_%s_%t.bak';

delete noprompt expired backup;

delete noprompt obsolete;

release channel c1;

release channel c2;

}

数据库增量备份

rman> run {

allocate channel c1 type disk;

allocate channel c2 type disk;

backup incremental level 1 database format '/home/oracle/backup/db_incre_%d_%s_%t.bak';

sql 'alter system archive log current';

crosscheck archivelog all;

backup archivelog all format '/home/oracle/backup/archlog_incre_%d_%s_%t.bak' delete input;

backup spfile format '/home/oracle/backup/spfile_%d_%s_%t.bak';

backup current controlfile format '/home/oracle/backup/controlfile_%d_%s_%t.bak';

delete noprompt expired backup;

delete noprompt obsolete;

release channel c1;

release channel c2;

}

RMAN数据库恢复

1. 数据文件丢失恢复

1. 找到丢失文件的数据文件号

1. 根据文件号查找数据备份列表 rman> list backup of datafile 4;
2. 将数据文件离线 rman> sql 'alter database datafile 4 offline'
3. 从备份文件将数据文件拷贝回来 rman> restore datafile 4;
4. 数据文件恢复到最新 rman> recover datafile 4;
5. 将数据文件上线rman> sql 'alter database datafile 4 online';

全部脚本如下：

rman> run {

sql 'alter database datafile 4 offline';

restore datafile 4;

recover datafile 4;

sql 'alter database datafile 4 online'

}

2. 全部数据文件丢失恢复

sql>shutdown immediate;

sql>startup mount;

rman> run {

restore database;

recover database;

sql 'alter database open';

}

3. 控制文件、数据文件丢失恢复

1. sql> startup nomount
2. 恢复控制文件rman> restore controlfile from '/home/oracle/backup/c_ORCL_*_*'
3. 恢复数据文件rman> restore database;
4. 介质恢复rman> run {

set until sequenct 63; //这里的63是存在的归档日志的最大的序列号

recover database;

}

5. 打开数据库,resetlogs会将redo日志序列号恢复从0开始，这样会导致之前的所有备份不再可用 ，可把之前的备份删除 rman> sql ‘alter database open resetlogs';

6. 把之前的备份删除 rman> delete backup;

7. 再给数据库做全备，参考上面RMAN数据库全备脚本

4. 数据库坏块恢复

1. 通过rman检查数据库上是否有坏块 rman> backup check logical validate database;
2. 查看数据库存在多少坏块 sql> select * from v$database_block_corruption;
3. 恢复一个坏块 rman> blockrecover datafile 4 block 20635;
4. 如果有多个坏块 rman> blockrecover corruption list;