Oracle 用户管理模式下的恢复案例-不完全恢复

1. 不完全恢复的几种常用方法

01. recover database using backup controlfile

        如果丢失当前控制文件，用冷备份的控制文件恢复的时候，用来告诉 oracle，不要以 controlfile 中的 scn 作为恢复的终点；

02. recover database until cancel

        如果丢失 current/active redo 的时候，手动指定终点。

recover database until cancel; --SQLPlus 使用 
recover database until time '2022-08-09:14:20:45' --SQLPlus 与 RMAN 都支持 
recover database unitl time '2022-08:14:20:45' using backup controlfile 
recover database until change 55555 --SQLPlus 使用 
recover database until scn 55555 --RMAN 使用 
recover database until sequence 20 --RMAN 使用 

03. recover database using backup controlfile until cancel;

        如果丢失当前 controlfile 并且 current/active redo 都丢失，会先去自动应用归档日志,可以实现最大的恢复；

04. recover database until cancel using backup controlfile;

        如果丢失当前 controlfile 并且 current/active redo 都丢失，以旧的 redo 中的 scn为恢复终点。因为没有应用归档日志，所以会丢失数据。

2. 实操 1 基于【until time】不完全恢复

说明：数据库需要在归档模式下面操作

        在数据库冷备份后，在表中插入一条数据，提交，记录当前数据库的时间，dorp掉表，删除数据库的dbf文件，控制文件和日志文件还在，然后把数据库冷备份拷贝过来，用recover database until time '2023-03-28 22:05:05';恢复数据库到插入数据之后，用 alter database open RESETLOGS;启动数据库，查询最后提交的数据没有丢失。

--先做冷备 
shutdown immediate; 
SQL> host rm -rf /backup/hfzcdb/*.* 
*/
SQL> host cp /oradata/hfzcdb/* /backup/hfzcdb/ --先做冷备 */ 
set time on; SYS@hfzcdb> shutdown immediate
SYS@hfzcdb> host cp /oradata/hfzcdb/* /backup/hfzcdb/
*/
SYS@hfzcdb> startup
ORACLE instance started.Total System Global Area 2466249672 bytes
Fixed Size                  8899528 bytes
Variable Size             671088640 bytes
Database Buffers         1778384896 bytes
Redo Buffers                7876608 bytes
Database mounted.
Database opened.
SYS@hfzcdb> set time on
22:02:35 SYS@hfzcdb> conn hfedu/hfedu123
Connected.
22:03:01 hfedu@hfzcdb> insert into hfedu3 values(44,'hfedu44');1 row created.22:03:22 hfedu@hfzcdb> commit；
22:04:38 hfedu@hfzcdb> select to_char(sysdate,'yyyy-mm-dd hh24:mi:ss') from dual;TO_CHAR(SYSDATE,' yyyy-mm-dd hh24:mi:ss')  
-------------------
2023-03-28 22:05:0522:05:05 hfedu@hfzcdb> drop table hfedu3;Table dropped.22:05:28 hfedu@hfzcdb> select * from hfedu3;
select * from hfedu3
*
ERROR at line 1:
ORA-00942: table or view does not exist
22:05:33 hfedu@hfzcdb> conn / as sysdba
Connected.
22:05:41 SYS@hfzcdb> shutdown immedaite
SP2-0717: illegal SHUTDOWN option
22:05:54 SYS@hfzcdb> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
22:06:34 SYS@hfzcdb> host rm /oradata/hfzcdb/*.dbf    
*/22:07:06 SYS@hfzcdb> host cp /backup/hfzcdb/*.dbf /oradata/hfzcdb/
*/
22:08:45 SYS@hfzcdb> startup mount
ORACLE instance started.Total System Global Area 2466249672 bytes
Fixed Size                  8899528 bytes
Variable Size             671088640 bytes
Database Buffers         1778384896 bytes
Redo Buffers                7876608 bytes
Database mounted.
22:09:11 SYS@hfzcdb> select file#,checkpoint_change# from v$datafile;FILE# CHECKPOINT_CHANGE#
---------- ------------------
1            2136506
2            2136506
3            2136506
4            2136506
5            2136506
6            21365066 rows selected.22:09:14 SYS@hfzcdb> select file#,checkpoint_change# from v$datafile_header;FILE# CHECKPOINT_CHANGE#
---------- ------------------
1            2135155
2            2135155
3            2135155
4            2135155
5            2135155
6            21351556 rows selected.22:09:33 SYS@hfzcdb> recover database until time '2023-03-28 22:05:05';
Media recovery complete.
22:10:40 SYS@hfzcdb> alter database open;
alter database open
*
ERROR at line 1:
ORA-01589: must use RESETLOGS or NORESETLOGS option for database open22:10:53 SYS@hfzcdb> alter database open RESETLOGS;Database altered.22:11:21 SYS@hfzcdb> select *from hfedu.hfedu3;ID NAME
---------- ----------------------------------------
1 hfedu01
2 hfedu02
3 hfedu03
4 hfedu04
5 hfedu5
6 hfedu6
44 hfedu447 rows selected.22:11:41 SYS@hfzcdb>

3. 实操 2 基于【Until Cancel】不完全恢复

        在数据库冷备份后，在表中插入一条数据，把数据刷到日志文件，在checkpoint到数据文件，再插入一条,没有刷到日志文件，删除/oradata/hfzcdb/下面所有的数据包括控制文件和日志文件，再把之前冷备份的数据拷贝回来，把数据库启动到nomount状态，重建控制文件，使用recover database using backup controlfile until cancel; 在原来的控制文件基础上面恢复，实现最大的恢复，选择自动恢复，然后再alter database open RESETLOGS;数据库，数据能恢复到插入的第一条数据，第二条没有写入归档日志的，恢复不了。

hfeduSQL> conn hfedu/hfedu123
Connected.
hfedu@hfzcdb> select *from hfedu;ID NAME
---------- ----------------------------------------
1 hfedu01
55 hfedu05
66 hfedu06
77 hfedu07
88 hfedu08hfedu@hfzcdb> insert into hfedu values(99,'hfedu09');1 row created.hfedu@hfzcdb> commit;Commit complete.hfedu@hfzcdb> alter system switch logfile;System altered.hfedu@hfzcdb> alter system checkpoint;System altered.
-----------------------
[oracle@hfzcdb91:/archive/HFZCDB/archivelog/2023_12_25]$ll -lst
total 92196
416 -rw-r----- 1 oracle oinstall   422400 Dec 25 16:10 o1_mf_1_1_lrlg7o4n_.arc
396 -rw-r----- 1 oracle oinstall   405504 Dec 25 16:01 o1_mf_1_1_lrlfpvlc_.arc
16 -rw-r----- 1 oracle oinstall    15872 Dec 25 15:46 o1_mf_1_3_lrldszlc_.arc
28 -rw-r----- 1 oracle oinstall    26112 Dec 25 15:43 o1_mf_1_2_lrldo70m_.arc
1512 -rw-r----- 1 oracle oinstall  1544704 Dec 25 15:38 o1_mf_1_1_lrldbxly_.arc
89828 -rw-r----- 1 oracle oinstall 91983872 Dec 25 14:51 o1_mf_1_19_lrl9n0wk_.arc
[oracle@hfzcdb91:/archive/HFZCDB/archivelog/2023_12_25]$strings o1_mf_1_1_lrlg7o4n_.arc |grep hfedu09
hfedu09
-----------------------
hfedu@hfzcdb> insert into hfedu values(991,'hfedu091');1 row created.hfedu@hfzcdb> commit;Commit complete.hfedu@hfzcdb> alter system checkpoint;System altered.hfedu@hfzcdb> host ll /oradata/hfzcdb/**/
/bin/bash: ll: command not foundhfedu@hfzcdb> host ll /oradata/hfzcdb/
/bin/bash: ll: command not foundhfedu@hfzcdb> host ls /oradata/hfzcdb/
control01.ctl  hfedu01.dbf  redo02.log  sysaux01.dbf  temp01.dbf     users01.dbf
control02.ctl  redo01.log   redo03.log  system01.dbf  undotbs01.dbfhfedu@hfzcdb> host rm -f /oradata/hfzcdb/*
*/
hfedu@hfzcdb> conn / as sysdba
Connected.
SYS@hfzcdb> shutdown abort
ORACLE instance shut down.
SYS@hfzcdb> exit
Disconnected from Oracle Database 19c Enterprise Edition Release 19.0.0.0.0 - Production
Version 19.3.0.0.0
exit[oracle@hfzcdb91:/home/oracle]$sqlplus / as sysdbaSQL*Plus: Release 19.0.0.0.0 - Production on Mon Dec 25 16:14:07 2023
Version 19.3.0.0.0Copyright (c) 1982, 2019, Oracle.  All rights reserved.Connected to an idle instance.hfeduSQL> host cp /backup/hfzcdb/*.dbf /oradata/hfzcdb/
*/
STARTUP NOMOUNT
CREATE CONTROLFILE REUSE DATABASE "HFZCDB" RESETLOGS  ARCHIVELOG
MAXLOGFILES 16
MAXLOGMEMBERS 3
MAXDATAFILES 8192
MAXINSTANCES 8
MAXLOGHISTORY 292
LOGFILE
GROUP 1 '/oradata/hfzcdb/redo01.log'  SIZE 200M BLOCKSIZE 512,
GROUP 2 '/oradata/hfzcdb/redo02.log'  SIZE 200M BLOCKSIZE 512,
GROUP 3 '/oradata/hfzcdb/redo03.log'  SIZE 200M BLOCKSIZE 512
-- STANDBY LOGFILE
DATAFILE
'/oradata/hfzcdb/system01.dbf',
'/oradata/hfzcdb/sysaux01.dbf',
'/oradata/hfzcdb/undotbs01.dbf',
'/oradata/hfzcdb/users01.dbf',
'/oradata/hfzcdb/hfedu01.dbf'
CHARACTER SET AL32UTF8
;
ORACLE instance started.Total System Global Area 2466249672 bytes
Fixed Size                  8899528 bytes
Variable Size             671088640 bytes
Database Buffers         1778384896 bytes
Redo Buffers                7876608 bytes
hfeduSQL>   2    3    4    5    6    7    8    9   10   11   12   13   14   15   16   17   18   19
Control file created.hfeduSQL>  recover database using backup controlfile until cancel;
ORA-00279: change 1428569 generated at 12/25/2023 15:49:24 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_12_25/o1_mf_1_4_%u_.arc
ORA-00280: change 1428569 for thread 1 is in sequence #4Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
auto
ORA-00283: recovery session canceled due to errors
ORA-19906: recovery target incarnation changed during recoveryORA-01112: media recovery not startedhfeduSQL>  recover database using backup controlfile until cancel;
ORA-00279: change 1428571 generated at 12/25/2023 16:05:05 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_12_25/o1_mf_1_1_lrlg7o4n_.arc
ORA-00280: change 1428571 for thread 1 is in sequence #1Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
/archive/HFZCDB/archivelog/2023_12_25/o1_mf_1_1_lrlg7o4n_.arc
ORA-00279: change 1429593 generated at 12/25/2023 16:10:29 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_12_25/o1_mf_1_2_%u_.arc
ORA-00280: change 1429593 for thread 1 is in sequence #2
ORA-00278: log file '/archive/HFZCDB/archivelog/2023_12_25/o1_mf_1_1_lrlg7o4n_.arc' no longer needed for this recoverySpecify log: {<RET>=suggested | filename | AUTO | CANCEL}
cancel
Media recovery cancelled.
hfeduSQL> alter database open resetlogs;Database altered.hfeduSQL> alter database open RESETLOGS;hfeduSQL> select * from hfedu.hfedu;ID NAME
---------- ----------------------------------------99 hfedu09
1 hfedu01
55 hfedu05
66 hfedu06
77 hfedu07
88 hfedu086 rows selected.hfeduSQL>

4. 实操 3 基于【until scn】不完全恢复

        在数据库冷备份后，插入一条数据提交，并写到日志文件，checkpoint到数据文件，记录此时scn，然后再插入一条数据提交，再删除一条数据，提交，关闭数据库，把数据文件恢复到冷备份的时候，启动数据库到nomount状态，使用 recover database until change 2182325;恢复数据库，再用 alter database open resetlogs;启动数据库后，查询数据，发现数据是插入第一条数据的状态。

--先做冷备 
shutdown immediate;
SQL> host rm -rf /backup/hfzcdb/*.* */
SQL> host cp /oradata/hfzcdb/* /backup/hfzcdb/ --先做冷备 */ 
set time on; hfeduSQL> conn hfedu/hfedu123
Connected.
hfedu@hfzcdb>
hfedu@hfzcdb> insert into hfedu3 values(66,'hfedu66');1 row created.hfedu@hfzcdb> commit
2  ;Commit complete.hfedu@hfzcdb> alter system switch logfile;System altered.hfedu@hfzcdb> /System altered.hfedu@hfzcdb> /
/System altered.hfedu@hfzcdb>
System altered.hfedu@hfzcdb> hfedu@hfzcdb> alter system checkpoint
2  ;System altered.hfedu@hfzcdb> select current_scn from v$database;CURRENT_SCN
-----------
2182325hfedu@hfzcdb> alter system checkpoint
2  ;System altered.hfedu@hfzcdb> alter system checkpoint;System altered.hfedu@hfzcdb> select current_scn from v$database;CURRENT_SCN
-----------
2182335hfedu@hfzcdb> alter system checkpoint;System altered.hfedu@hfzcdb> select current_scn from v$database;CURRENT_SCN
-----------
2182342hfedu@hfzcdb> insert into hfedu3 values(77,'hfedu77');1 row created.hfedu@hfzcdb> commit
2  ;Commit complete.hfedu@hfzcdb> delete from hfedu where id=1;
delete from hfedu where id=1
*
ERROR at line 1:
ORA-00942: table or view does not existhfedu@hfzcdb> delete from hfedu3 where id=1;1 row deleted.hfedu@hfzcdb> commit
2  ;Commit complete.hfedu@hfzcdb> alter system checkpoint;System altered.hfedu@hfzcdb> select current_scn from v$database;CURRENT_SCN
-----------
2182385hfedu@hfzcdb> select * from hfedu.hfedu3;ID NAME
---------- ----------------------------------------
66 hfedu66
77 hfedu77
2 hfedu02
3 hfedu03
4 hfedu04
5 hfedu5
6 hfedu6
44 hfedu448 rows selected.hfedu@hfzcdb> conn / as sysdba
Connected.
SYS@hfzcdb> shutdown abort
ORACLE instance shut down.
SYS@hfzcdb> host cp /backup/hfzcdb/*.dbf /oradata/hfzcdb/   【覆盖现在日志，就和删除恢复一样】
*/
SYS@hfzcdb> startup nomount
ORACLE instance started.Total System Global Area 2466249672 bytes
Fixed Size                  8899528 bytes
Variable Size             671088640 bytes
Database Buffers         1778384896 bytes
Redo Buffers                7876608 bytes
SYS@hfzcdb> alter database mount
2  ;Database altered.SYS@hfzcdb> recover database until change 2182325;
ORA-00279: change 2135155 generated at 03/28/2023 22:00:32 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_03_28/o1_mf_1_35_l25xd0yp_.arc
ORA-00280: change 2135155 for thread 1 is in sequence #35Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
auto
ORA-00279: change 2136023 generated at 03/28/2023 22:11:12 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_03_29/o1_mf_1_1_l27n4015_.arc
ORA-00280: change 2136023 for thread 1 is in sequence #1
ORA-00278: log file '/archive/HFZCDB/archivelog/2023_03_28/o1_mf_1_35_l25xd0yp_.arc' no longer needed for this recoveryORA-00279: change 2177944 generated at 03/29/2023 15:15:29 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_03_29/o1_mf_1_1_l27wtv68_.arc
ORA-00280: change 2177944 for thread 1 is in sequence #1ORA-00279: change 2182271 generated at 03/29/2023 16:14:19 needed for thread 1
ORA-00289: suggestion : /archive/HFZCDB/archivelog/2023_03_29/o1_mf_1_2_l27wv3w1_.arc
ORA-00280: change 2182271 for thread 1 is in sequence #2Log applied.
Media recovery complete.
SYS@hfzcdb> alter database open resetlogs;  【数据库只要是不完全恢复，都需要resetlogs】Database altered.SYS@hfzcdb> select * from hfedu.hfedu3;ID NAME
---------- ----------------------------------------
66 hfedu66
1 hfedu01
2 hfedu02
3 hfedu03
4 hfedu04
5 hfedu5
6 hfedu6
44 hfedu448 rows selected.SYS@hfzcdb>