测试oracle删除全库之后使用rman恢复

说明：项目生产环境所使用的rman备份脚本使用的是incremental level 0进行全库备份。这里是使用了生产环境所用的备份脚本进行测试。1-3为测试步骤：

1. backup database use rman

oracle@train_ocsdb[/]$ rman target /

Recovery Manager: Release 12.2.0.1.0 - Production on Mon Jan 6 17:29:00 2020

Copyright (c) 1982, 2017, Oracle and/or its affiliates.  All rights reserved.

connected to target database: CC (DBID=2213285106)

RMAN> CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '/ccbackup/cf_%F';

RMAN> run {

allocate channel 'dev1' type disk;

allocate channel 'dev2' type disk;

allocate channel 'dev3' type disk;

allocate channel 'dev4' type disk;

backup incremental level 0 tag 'dbL0' format '/ccbackup/%d_Incr0_%T_%s_%u' database INCLUDE CURRENT CONTROLFILE;

backup format '/ccbackup/%d_log_%T_%U' archivelog until time 'sysdate' delete all input ;

release channel 'dev1';

release channel 'dev2';

release channel 'dev3';

release channel 'dev4';

delete noprompt obsolete;

}

2. drop database

SQL> shutdown immediate;

Database closed.

Database dismounted.

ORACLE instance shut down.

SQL> startup mount;

SQL> alter system enable restricted session;

SQL> drop database;

3. recover database

1. database started to nomount
   RMAN> startup nomount
2. Restore spfile from backup
   RMAN> restore spfile from '/archcc/fast_recovery_area/CC/autobackup/2020_01_07/o1_mf_s_1029092020_h19zh4nw_.bkp';
   RMAN> shutdown immediate;
   RMAN> startup nomount
3. Restore control file from backup
   RMAN> restore controlfile from '/ccbackup/cf_c-2213285106-20200107-04';
   RMAN> alter database mount;
4. Setting recovery catalog
   RMAN> catalog start with '/ccbackup/';
5. RMAN restore and recover database
   RMAN>run {
   allocate channel 'd1' type disk;
   allocate channel 'd2' type disk;
   allocate channel 'd3' type disk;
   allocate channel 'd4' type disk;
   restore database;
   recover database;
   release channel d1;
   release channel d2;
   release channel d3;
   release channel d4;
   }
   6.open database
   RMAN> alter database open resetlogs;

4. 问题描述

在recover database时报错如下：

这是因为控制文件太旧而导致的报错，找不到归档日志。

一般情况下，recover database都会出现以上的报错，因为归档的备份没有记录在控制文件中，alter database open resetlogs 可以打开数据库，而在incremental 0全库备份的情况下，alter database open resetlogs 会提示报错。

5. 问题分析

1. 分析备份脚本

backup incremental level 0 tag 'dbL0' format '/ccbackup/%d_Incr0_%T_%s_%u' database INCLUDE CURRENT CONTROLFILE;--- 全库备份，也备份控制文件

backup format '/ccbackup/%d_log_%T_%U' archivelog until time 'sysdate' delete all input ; ---删除现有的归档日志，然后备份新产生的归档日志。

2. 查看备份文件

由于测试rman备份期间，数据库没有任何操作，所以没有产生归档日志，rman只备份了数据文件和控制文件，没有备份归档日志。进而在恢复时会报错。

3. rman恢复步骤

首先恢复spfile，然后恢复controlfile，然后根据controlfile的信息来恢复数据文件，最后根据归档来恢复controlfile没有记录到的数据变化。

incremental 0的备份方式在恢复全库时，会去查找备份的归档日志，如果不存在，则会报错。

由于在测试时，数据库没有任何操作，没有产生新的归档日志，所以在alter database open时报错。

6. 问题解决

1. 在rman备份期间，手动切归档：alter system switch logfile
2. 更改脚本

run {

allocate channel 'dev1' type disk;

allocate channel 'dev2' type disk;

allocate channel 'dev3' type disk;

allocate channel 'dev4' type disk;

backup incremental level 0 format '/oradata/export/rman/%d_Incr0%T%s_%u' database;

sql 'alter system archive log current';

backup format '/oradata/export/rman/%d_log%T%U' archivelog until time 'sysdate' delete all input ;

release channel 'dev1';

release channel 'dev2';

release channel 'dev3';

release channel 'dev4';

delete noprompt obsolete;

}

1. 总结：

1） 在生产环境中，数据库会产生归档日志，所以不会出现恢复时报错的问题。

2） backup database和incremetal 0虽然都是备份全库的命令，但是在恢复时backup database即使没有归档也能恢复，incremetal 0则必须要有归档日志。