达梦8备份与恢复
DM8物理备份与恢复
目录
DM8物理备份与恢复
1、联机完全备份与恢复
2、脱机完全备份与恢复
3、联机增量备份与恢复
-
物理备份：直接扫描数据库文件，找出那些已经分配、使用的数据页，拷贝并保存到备份集中。物理备份过程中，不关心数据页的具体内容是什么，也不关心数据页属于哪一张表，只是简单的根据数据库文件系统的描述，来挑选有效的数据页。
-
物理备份根据数据库服务状态分为：联机备份（热备），脱机备份（冷备）。
-
根据备份策略分为：完全备份和增量备份。
-
联机备份：数据库服务运行的情况下进行备份。
-
脱机备份：数据库服务关闭的情况下进行备份。
-
完全备份：执行完全备份，备份程序会扫描数据文件，拷贝所有被分配、使用的数据页，写入到备份片文件中库备份会扫描整个数据库的所有数据文件。
-
增量备份：在某个特定备份集基础上，收集数据库新修改的数据页进行备份，可以有效 减少备份集的空间占用、提高备份速度。
-
个人总结：无论是联机备份还是脱机备份，只要是库级的物理备份，恢复的时候都需要脱机进行；
完全备份只需一条语句，脱机则要分三部曲，restore-->recover-->recover（三步在下面的实验中有演示）;
增量备份是在其他备份集基础上做的备份，恢复的时候也要在其他备份集恢复之后再做恢复，增量备份的恢复语 句和完全备份的恢复语法结构完全一致。
1、联机完全备份与恢复
步骤：开启归档-->联机备份-->删除数据-->关闭数据库服务-->执行三个恢复语句进行数据恢复-->启动数据库服务验证数据是否正确
1、开启归档模式：
正在上传…重新上传取消
1）查看是否数据库是否处于归档状态（N为非归档状态，Y为归档状态）
SQL> select arch_mode from v$database;
LINEID ARCH_MODE
---------- ---------
1 N
used time: 10.513(ms). Execute id is 52.2）修改数据库为配置状态
SQL> alter database mount;
executed successfully
used time: 00:00:01.769. Execute id is 0.3）添加归档类型及目录并开启归档
SQL> alter database add archivelog 'DEST=/dm8/arch,type=local';
executed successfully
used time: 36.225(ms). Execute id is 0.
SQL> alter database archivelog;
executed successfully
used time: 5.734(ms). Execute id is 0.4）修改数据库为open状态
SQL> alter database open;
executed successfully
used time: 00:00:01.799. Execute id is 0.
SQL> select arch_mode from v$database;
LINEID ARCH_MODE
---------- ---------
1 Y
used time: 1.047(ms). Execute id is 53.
2、开始备份
1）初始数据：
SQL> select * from test1;
LINEID ID
---------- -----------
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
10 rows got
used time: 2.101(ms). Execute id is 54.2）开始联机备份：指定备份集路径为：/dm8/data/full_bak
SQL> BACKUP DATABASE BACKUPSET '/dm8/data/full_bak';
executed successfully
used time: 00:00:02.112. Execute id is 55.3）备份成功，删除表：
SQL> drop table test1;
executed successfully
used time: 128.207(ms). Execute id is 56.
SQL> commit;
executed successfully
used time: 0.490(ms). Execute id is 57.
4）关闭数据库服务恢复数据（无论联机备份还是脱机备份，在恢复的时候都需要脱机恢复）
[dmdba@localhost bin]$ ./DmServiceDMSERVER stop
Stopping DmServiceDMSERVER: [ OK ]5）开始恢复数据：
还原数据库：数据库的还原包括数据库配置文件还原和数据文件还原，通过 RESTORE 命令还原后的数据库不可用，需进一步执行 RECOVER 命令，将数据库恢复到备份结束时的状态。
[dmdba@localhost bin]$ ./dmrman
dmrman V8
RMAN> restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
file dm.key not found, use default license!
RESTORE DATABASE CHECK......
RESTORE DATABASE,data collect......
RESTORE DATABASE,database refresh ......
RESTORE BACKUPSET [/dm8/data/full_bak] START......
total 5 packages processed...
total 6 packages processed...
RESTORE DATABASE,UPDATE ctl file......
RESTORE DATABASE,REBUILD key file......
RESTORE DATABASE,CHECK db info......
RESTORE DATABASE,UPDATE db info......
total 6 packages processed...
total 6 packages processed!
CMD END.CODE:[0]
restore successfully.
time used: 00:00:01.1956）恢复数据库
RMAN> recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[43462]
RESTORE RLOG CHECK......
RESTORE RLOG, gen tmp file......
RESTORE RLOG FROM BACKUPSET [/dm8/data/full_bak] START......
total 2 packages processed...
total 2 packages processed...
total 2 packages processed!
CMD END.CODE:[0]
EP:0 total 1 pkgs applied, percent: 50%
EP:0 total 2 pkgs applied, percent: 100%
Recover from archive log finished, time used:0.017s.
recover successfully!
time used: 275.419(ms)
7）更新数据库
数据库更新是指更新数据库的 DB_MAGIC，并将数据库调整为可正常工作状态。数据库更新发生在重做 REDO 日志恢复数据库后，或者目标库不需要执行重做日志已经处于一致状态的情况。
RMAN> recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[43466]
EP[0]'s apply_lsn[43466] >= end_lsn[43466]
recover successfully!
time used: 00:00:01.0098）恢复过程中没有报错，启动数据库服务，查询数据并对比没有错误。
[dmdba@localhost bin]$ ./DmServiceDMSERVER start
Starting DmServiceDMSERVER: [ OK ]
[dmdba@localhost bin]$ ./disql
disql V8
username:
password:
Server[LOCALHOST:5236]:mode is normal, state is open
login used time: 7.544(ms)
SQL> select * from test1;
LINEID ID
---------- -----------
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
10 rows got
used time: 12.037(ms). Execute id is 4.
2、脱机完全备份与恢复
关闭数据库服务的情况下使用rman工具（rman工具为数据库自带工具）进行备份：
步骤：关闭数据库服务-->脱机备份-->启动数据库服务并删除数据（也可以直接在一个新初始化的实例中做恢复）-->关闭数据库服务-->执行三个恢复语句-->启动服务并验证是否恢复成功：
和前面的联机备份恢复一样，先备份，再删除一些数据，然后进行恢复，看是否成功：
1）停掉数据库服务并使用数据库安装目录下的bin目录下的rman工具进行完全备份：
[dmdba@localhost bin]$ ./DmServiceDMSERVER stop
Stopping DmServiceDMSERVER: [ OK ]
[dmdba@localhost bin]$ ./dmrman CTLSTMT="BACKUP DATABASE '/dm8/data/DAMENG/dm.ini' FULL TO BACKUP_FILE1 BACKUPSET '/dm8/data/BACKUP_FILE_01'"
dmrman V8
BACKUP DATABASE '/dm8/data/DAMENG/dm.ini' FULL TO BACKUP_FILE1 BACKUPSET '/dm8/data/BACKUP_FILE_01'
file dm.key not found, use default license!
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[46019]
BACKUP DATABASE [DAMENG],execute......
CMD CHECK LSN......
BACKUP DATABASE [DAMENG],collect dbf......
CMD CHECK ......
DBF BACKUP SUBS......
total 1 packages processed...
total 2 packages processed...
total 3 packages processed...
total 4 packages processed...
total 5 packages processed...
DBF BACKUP MAIN......
BACKUPSET [/dm8/data/BACKUP_FILE_01] END, CODE [0]......
META GENERATING......
total 6 packages processed...
total 6 packages processed...
total 6 packages processed!
CMD END.CODE:[0]
backup successfully!
time used: 00:00:01.3112）删除数据，和上面的联机备份操作一样此处省略；
3）还原数据库restore（和上面联机备份一样）
[dmdba@localhost bin]$ ./dmrman CTLSTMT="RESTORE DATABASE '/dm8/data/DAMENG/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01'"
dmrman V8
RESTORE DATABASE '/dm8/data/DAMENG/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01'
file dm.key not found, use default license!
RESTORE DATABASE CHECK......
RESTORE DATABASE,data collect......
RESTORE DATABASE,database refresh ......
RESTORE BACKUPSET [/dm8/data/BACKUP_FILE_01] START......
total 5 packages processed...
total 6 packages processed...
RESTORE DATABASE,UPDATE ctl file......
RESTORE DATABASE,REBUILD key file......
RESTORE DATABASE,CHECK db info......
RESTORE DATABASE,UPDATE db info......
total 6 packages processed...
total 6 packages processed!
CMD END.CODE:[0]
restore successfully.
time used: 971.740(ms)4）恢复数据库（recover）
[dmdba@localhost bin]$ ./dmrman CTLSTMT="RECOVER DATABASE '/dm8/data/DAMENG/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01'"
dmrman V8
RECOVER DATABASE '/dm8/data/DAMENG/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01'
file dm.key not found, use default license!
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[46019]
RESTORE RLOG CHECK......
CMD END.CODE:[603],DESC:[no log generates while the backupset [/dm8/data/BACKUP_FILE_01] created]
no log generates while the backupset [/dm8/data/BACKUP_FILE_01] created
recover successfully!
time used: 250.328(ms)5）更新数据库（UPDATE DB_MAGIC）
[dmdba@localhost bin]$ ./dmrman CTLSTMT="RECOVER DATABASE '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC"
dmrman V8
RECOVER DATABASE '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC
file dm.key not found, use default license!
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[46019]
EP[0]'s apply_lsn[46019] >= end_lsn[46019]
recover successfully!
time used: 00:00:01.009
3、联机增量备份与恢复
增量备份是在某个特定备份集基础上，收集数据库新修改的数据页进行备份，可以有效减少备份集的空间占用、提高备份速度。
与全量备份一样，增量备份可以脱机备份也可以联机备份，以联机备份为例：
步骤：全备-->增加或删除数据-->增量备份-->删除所有的数据（或者在一个新初始化的实例上做备份）并关闭数据库服务-->全量备份恢复-->增量备份恢复-->启动数据库服务验证数据是否无误
1）执行全备
SQL> BACKUP DATABASE BACKUPSET '/dm8/data/full_bak';
executed successfully
used time: 00:00:01.451. Execute id is 13.2）修改数据库中的数据：
SQL> create table test3 as select * from test1 where id<5;
executed successfully
used time: 12.606(ms). Execute id is 14.3）执行增备：
SQL> BACKUP DATABASE INCREMENT WITH BACKUPDIR '/dm8/data/full_bak' BACKUPSET '/dm8/data/full_bak/db_increment_bak_02';
executed successfully
used time: 856.383(ms). Execute id is 16.4）删除数据库中的test1和test3表以后进行恢复；
5） 先恢复完全备份（恢复方式和前面的联机备份恢复方法一样，这里每一步不再做详细解释）
[dmdba@localhost bin]$ ./dmrman
dmrman V8
RMAN> restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
file dm.key not found, use default license!
RESTORE DATABASE CHECK......
RESTORE DATABASE,data collect......
RESTORE DATABASE,database refresh ......
RESTORE BACKUPSET [/dm8/data/full_bak] START......
total 5 packages processed...
total 6 packages processed...
RESTORE DATABASE,UPDATE ctl file......
RESTORE DATABASE,REBUILD key file......
RESTORE DATABASE,CHECK db info......
RESTORE DATABASE,UPDATE db info......
total 6 packages processed...
total 6 packages processed!
CMD END.CODE:[0]
restore successfully.
time used: 00:00:01.193
RMAN> recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/';
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[47313]
RESTORE RLOG CHECK......
RESTORE RLOG, gen tmp file......
RESTORE RLOG FROM BACKUPSET [/dm8/data/full_bak] START......
total 2 packages processed...
total 2 packages processed...
total 2 packages processed!
CMD END.CODE:[0]
EP:0 total 1 pkgs applied, percent: 14%
EP:0 total 2 pkgs applied, percent: 28%
EP:0 total 3 pkgs applied, percent: 42%
EP:0 total 4 pkgs applied, percent: 57%
EP:0 total 5 pkgs applied, percent: 71%
EP:0 total 6 pkgs applied, percent: 85%
EP:0 total 7 pkgs applied, percent: 100%
Recover from archive log finished, time used:0.010s.
recover successfully!
time used: 280.686(ms)
RMAN> recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[47331]
EP[0]'s apply_lsn[47331] >= end_lsn[47331]
recover successfully!
time used: 00:00:01.0086）在完全备份恢复的基础上做增量恢复（也就是再做一次跟前面的恢复完全一样的操作，没有其他特殊操作）
RMAN> restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/db_increment_bak_02';
restore database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/db_increment_bak_02';
RESTORE DATABASE CHECK......
RESTORE DATABASE,data collect......
RESTORE DATABASE,database refresh ......
RESTORE BACKUPSET [/dm8/data/full_bak] START......
total 5 packages processed...
total 6 packages processed...
RESTORE BACKUPSET [/dm8/data/full_bak/db_increment_bak_02] START......
total 11 packages processed...
total 12 packages processed...
RESTORE DATABASE,UPDATE ctl file......
RESTORE DATABASE,REBUILD key file......
RESTORE DATABASE,CHECK db info......
RESTORE DATABASE,UPDATE db info......
total 12 packages processed...
total 12 packages processed!
CMD END.CODE:[0]
restore successfully.
time used: 00:00:01.165
RMAN> recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/db_increment_bak_02';
recover database '/dm8/data/DAMENG/dm.ini' from backupset '/dm8/data/full_bak/db_increment_bak_02';
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[47364]
RESTORE RLOG CHECK......
RESTORE RLOG, gen tmp file......
RESTORE RLOG FROM BACKUPSET [/dm8/data/full_bak/db_increment_bak_02] START......
total 2 packages processed...
total 2 packages processed...
total 2 packages processed!
CMD END.CODE:[0]
EP:0 total 1 pkgs applied, percent: 50%
EP:0 total 2 pkgs applied, percent: 100%
Recover from archive log finished, time used:0.005s.
recover successfully!
time used: 269.534(ms)
RMAN> recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
recover database '/dm8/data/DAMENG/dm.ini' UPDATE DB_MAGIC;
Database mode = 0, oguid = 0
EP[0]'s cur_lsn[47367]
EP[0]'s apply_lsn[47367] >= end_lsn[47367]
recover successfully!
time used: 995.346(ms)第一次发表博客，不足之处请多指教。