Oracle Study案例之--基于表空间的时间点恢复(TSPITR)
TSPITR(表空间时间点恢复)用于将一个或多个表空间恢复到过去某个时间点的状态,而其他表空间仍然保持现有状态。
TSPITR 相关的概念和术语:
(1) TSPITR (Tablespace Point-In-Time Recover)。TSPITR 是表空间时间点恢复的英文缩写格式,它表示将一个或多个表空间恢复到过去时间点的状态,而其他表空间仍然保持现有状态。
(2) TSPITR 实现方法。当实现表空间时间点恢复时,既可以使用用户管理的表空间时间点恢复方法,也可以使用RMAN 管理的表空间时间点恢复。
(3) DBPITR (Database Point-In-Time Recovery)。DBPITR 是数据库时间点恢复的英文缩写格式,它表示将数据库的所有表空间恢复到过去时间点的状态。注意,DBPITR 只适用于ARCHIVELOG 模式。
(4) 主数据库(Primary Database)。主数据库是指用于存放应用系统数据的Oracle 数据库,也被称为产品数据库或目标数据库。当执行TSPITR 时,主数据库是指包含被恢复表空间的数据库。
(5) 恢复集(Recovery Set)。恢复集是指在主数据库上需要执行 TSPITR 的表空间集合。注意,当在恢复集的表空间上执行TSPITR 时,要求这些表空间必须是自包含的。
(6) 辅助数据库(Auxiliary Database)。辅助数据库是主数据库的一个副本数据库。当执行TSPITR 时,辅助数据库用于将恢复集表空间恢复到过去时间点。注意,辅助数据库的所有物理文件都是从主数据库备份中取得,并且辅助数据库必须包含SYSTEM 表空间、UNDO 表空间以及恢复集表空间的备份文件。
(7) 辅助集(Auxiliary Set)。辅助集是指辅助数据库所需要的、除了恢复集表空间文件之外的其他文件集合。当执行 TSPITR 时,辅助数据库除了需要恢复集表空间的备份文件之外,还需要备份控制文件、SYSTEM 表空间的备份文件、UNDO 表空间的备分文件。
一、TSPITR相关基本概念
什么是TSPITR?
TSPITR实际上是一种时间点恢复,只不过这里的恢复是针对单个表空间而言的,利用这种方法可以将数据库中的某一个或几个非系统表空间恢复到过去的某个时刻而保持其他的表空间不变。
Oracle只有在控制文件、数据文件头、联机日志文件三者的scn都一致的情况才能正常打开数据库,所以我们不可能在需要进行TSPITR的数据库中直接进行恢复,这就牵涉到另外一个概念辅助实例(auxiliary instance)/辅助数据库(auxiliary database)
主数据库、辅助数据库、辅助实例、恢复集、辅助集
主数据库是指需要进行TSPITR的数据库。 辅助数据库是主数据库的一个副本或是其一个子集,用于进行TSPITR,因为TSPITR不能直接在主数据库上进行。 辅助实例对应于辅助数据库的一个实例。 恢复集是指构成需要进行TSPITR的表空间的所有数据文件的备份。 辅助集是指表空间进行恢复所需的其它所有文件,包括system\undo\temp 表空间的备份文件、备份控制文件、归档日志文件、口令文件、参数文件等。
案例分析:基于表空间的时间点恢复
1、测试环境 03:57:50 SQL> conn scott/tiger Connected. 03:57:57 SQL> 03:59:43 SQL> select * from tab; TNAME TABTYPE CLUSTERID ------------------------------ ------- ---------- EXCEPTIONS TABLE LXTB1 TABLE LXTB2 TABLE SALGRADE TABLE SYS_TEMP_FBT TABLE BONUS TABLE DEPT TABLE EMP TABLE 8 rows selected. 03:59:45 SQL> select table_name,tablespace_name from user_tables; TABLE_NAME TABLESPACE_NAME ------------------------------ ------------------------------ EXCEPTIONS USERS LXTB1 USERS LXTB2 LXTBS2 SALGRADE USERS BONUS USERS DEPT USERS EMP USERS SYS_TEMP_FBT 8 rows selected. 04:00:08 SQL> select * from lxtb2; ID ---------- 1 2 3 4 5 6 7 8 10 11 10 rows selected. 2、对主库做热备份 04:00:09 SQL> conn /as sysdba Connected. 04:00:14 SQL> 04:00:14 SQL> @/home/oracle/test_hot_bak 04:00:22 SQL> set feedback off pagesize 0 heading off verify off linesize 100 trimspool on echo off time off ***spooling to /disk1/backup/test/hot_cmd.sql spool /disk1/backup/test/hot_bak/hot_bak.lst alter system switch logfile; alter tablespace SYSTEM begin backup; host cp /u01/app/oracle/oradata/test/system01.dbf /disk1/backup/test/hot_bak alter tablespace SYSTEM end backup; alter tablespace RTBS begin backup; host cp /u01/app/oracle/oradata/test/rtbs01.dbf /disk1/backup/test/hot_bak alter tablespace RTBS end backup; alter tablespace SYSAUX begin backup; host cp /u01/app/oracle/oradata/test/sysaux01.dbf /disk1/backup/test/hot_bak alter tablespace SYSAUX end backup; alter tablespace USERS begin backup; host cp /u01/app/oracle/oradata/test/users01.dbf /disk1/backup/test/hot_bak alter tablespace USERS end backup; alter tablespace LOB_16K begin backup; host cp /u01/app/oracle/oradata/test/lob_16k01.dbf /disk1/backup/test/hot_bak alter tablespace LOB_16K end backup; alter tablespace UNDOTBS1 begin backup; host cp /u01/app/oracle/oradata/test/undotbs1.dbf /disk1/backup/test/hot_bak alter tablespace UNDOTBS1 end backup; alter tablespace INDX begin backup; host cp /u01/app/oracle/oradata/test/indx01.dbf /disk1/backup/test/hot_bak alter tablespace INDX end backup; alter tablespace LXTBS1 begin backup; host cp /u01/app/oracle/oradata/test/lxtbs01.dbf /disk1/backup/test/hot_bak alter tablespace LXTBS1 end backup; alter tablespace LXTBS2 begin backup; host cp /u01/app/oracle/oradata/test/lxtbs2.dbf /disk1/backup/test/hot_bak alter tablespace LXTBS2 end backup; archive log list; spool off; Database log mode Archive Mode Automatic archival Enabled Archive destination /disk4/arch/test Oldest online log sequence 6 Next log sequence to archive 9 Current log sequence 9 3、备份控制文件,作为备库的controlfile SQL> alter database backup controlfile to '/u01/app/oracle/oradata/test2/control01.ctl'; SQL> exit Disconnected from Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production With the Partitioning, OLAP and Data Mining options 4、生成pfile 文件 SQL> create pfile from spfile; 5、表被误操作(truncate) [oracle@work test]$ sqlplus / as sysdba; SQL*Plus: Release 10.2.0.1.0 - Production on Thu Oct 27 04:01:49 2011 Copyright (c) 1982, 2005, Oracle. All rights reserved. Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production With the Partitioning, OLAP and Data Mining options 04:01:50 SQL> 04:01:50 SQL> insert into scott.lxtb2 values (9); 1 row created. 04:02:15 SQL> insert into scott.lxtb2 values (12); 1 row created. 04:02:17 SQL> insert into scott.lxtb2 values (13); 1 row created. 04:02:19 SQL> commit; Commit complete. 04:02:21 SQL> alter system switch logfile; System altered. 04:02:39 SQL> / System altered. 04:02:40 SQL> / System altered. 04:02:41 SQL> truncate table scott.lxtb2; Table truncated. 04:02:50 SQL> select * from scott.lxtb2; no rows selected 04:02:56 SQL> ! 6、生成备库的pfile 文件(inittest2.ora)和备库口令文件 [oracle@work dbs]$ orapwd file=orapwtest2 password=oracle entries=3 force=y [oracle@work hot_bak]$ cp $ORACLE_HOME/dbs/inittest.ora $ORACLE_HOME/dbs/inittest2.ora --------备库pfile 文件 [oracle@work dbs]$ cat inittest2.ora test.__db_cache_size=251658240 test.__java_pool_size=4194304 test.__large_pool_size=4194304 test.__shared_pool_size=150994944 test.__streams_pool_size=0 *.audit_trail='NONE' *.background_dump_dest='$ORACLE_BASE/admin/test/bdump' *.core_dump_dest='$ORACLE_BASE/admin/test/cdump' *.db_16k_cache_size=12582912 *.db_4k_cache_size=12582912 *.db_block_size=8192 *.db_cache_size=30M#DEMO *.db_file_multiblock_read_count=16 *.db_name='test' *.db_recovery_file_dest='/disk1/flash/test' *.db_recovery_file_dest_size=2147483648 *.fast_start_mttr_target=900 *.fast_start_parallel_rollback='HIGH' *.log_archive_dest_1='location=/disk4/arch/test' *.log_archive_dest_2='' *.log_archive_format='arch_%t_%s_%r.log' *.log_checkpoints_to_alert=TRUE *.nls_date_format='yyyy-mm-dd hh24:mi:ss' *.O7_DICTIONARY_ACCESSIBILITY=FALSE *.optimizer_mode='choose' *.parallel_threads_per_cpu=4#SMALL *.pga_aggregate_target=10485760 *.query_rewrite_enabled='true' *.query_rewrite_integrity='trusted' *.recovery_parallelism=4 *.remote_login_passwordfile='EXCLUSIVE' *.resource_limit=TRUE *.sga_max_size=440401920 *.sga_target=418m *.shared_pool_size=100M#DEMO *.star_transformation_enabled='true' *.undo_management='auto' *.undo_tablespace='UNDOTBS1' *.user_dump_dest='$ORACLE_BASE/admin/test/udump' *.utl_file_dir='/home/oracle/logmnr' ---------添加以下内容 *.control_files='/u01/app/oracle/oradata/test2/control01.ctl' db_unique_name = test2 db_file_name_convert=('/u01/app/oracle/oradata/test','/u01/app/oracle/oradata/test2') log_file_name_convert=('/u01/app/oracle/oradata/test','/u01/app/oracle/oradata/test2','/disk1/oradata/test','/disk1/oradata/test2') 7、拷贝主库备份数据文件到备库目录下 [oracle@work hot_bak]$ mkdir /u01/app/oracle/oradata/test2 [oracle@work hot_bak]$ mkdir /disk1/oradata/test2 [oracle@work test]$ cd /disk1/backup/test/hot_bak/ [oracle@work hot_bak]$ ls hot_bak.lst indx01.dbf lob_16k01.dbf lxtbs01.dbf lxtbs2.dbf rtbs01.dbf sysaux01.dbf system01.dbf undotbs1.dbf users01.dbf [oracle@work hot_bak]$ cp *.dbf /u01/app/oracle/oradata/test2/ 8、启动备库到nomount export ORACLE_SID=test2 [oracle@work hot_bak]$ export ORACLE_SID=test2 [oracle@work hot_bak]$ !sql sqlplus / as sysdba; SQL*Plus: Release 10.2.0.1.0 - Production on Thu Oct 27 04:05:28 2011 Copyright (c) 1982, 2005, Oracle. All rights reserved. Connected to an idle instance. 04:05:29 SQL> 04:05:29 SQL> startup nomount pfile=$ORACLE_HOME/dbs/inittest2.ora ORACLE instance started. Total System Global Area 440401920 bytes Fixed Size 1219904 bytes Variable Size 113246912 bytes Database Buffers 322961408 bytes Redo Buffers 2973696 bytes 8、启动备库到mount状态 04:05:43 SQL> alter database mount clone database; Database altered. 04:06:25 SQL> col name for a50 04:06:32 SQL> select name ,file# ,status from v$datafile; NAME FILE# STATUS -------------------------------------------------- ---------- ------- /u01/app/oracle/oradata/test2/system01.dbf 1 SYSOFF /u01/app/oracle/oradata/test2/rtbs01.dbf 2 OFFLINE /u01/app/oracle/oradata/test2/sysaux01.dbf 3 OFFLINE /u01/app/oracle/oradata/test2/users01.dbf 4 OFFLINE /u01/app/oracle/oradata/test2/lob_16k01.dbf 5 OFFLINE /u01/app/oracle/oradata/test2/lxtbs01.dbf 6 OFFLINE /u01/app/oracle/oradata/test2/lxtbs2.dbf 7 OFFLINE /u01/app/oracle/oradata/test2/undotbs1.dbf 9 OFFLINE /u01/app/oracle/oradata/test2/indx01.dbf 14 OFFLINE 9 rows selected. 10、将数据文件联机 04:09:17 SQL> alter database datafile 1 online; Database altered. 04:09:24 SQL> alter database datafile 2 online; Database altered. 04:09:26 SQL> alter database datafile 3 online; Database altered. 04:09:28 SQL> alter database datafile 4 online; Database altered. 04:09:30 SQL> alter database datafile 5 online; Database altered. 04:09:32 SQL> alter database datafile 6 online; Database altered. 04:09:34 SQL> alter database datafile 7 online; Database altered. 04:09:36 SQL> alter database datafile 9 online; Database altered. 04:09:40 SQL> alter database datafile 14 online; Database altered. 04:09:42 SQL> select name ,file# ,status from v$datafile 04:09:47 2 ; NAME FILE# STATUS -------------------------------------------------- ---------- ------- /u01/app/oracle/oradata/test2/system01.dbf 1 SYSTEM /u01/app/oracle/oradata/test2/rtbs01.dbf 2 ONLINE /u01/app/oracle/oradata/test2/sysaux01.dbf 3 ONLINE /u01/app/oracle/oradata/test2/users01.dbf 4 ONLINE /u01/app/oracle/oradata/test2/lob_16k01.dbf 5 ONLINE /u01/app/oracle/oradata/test2/lxtbs01.dbf 6 ONLINE /u01/app/oracle/oradata/test2/lxtbs2.dbf 7 ONLINE /u01/app/oracle/oradata/test2/undotbs1.dbf 9 ONLINE /u01/app/oracle/oradata/test2/indx01.dbf 14 ONLINE 9 rows selected. 11、在备库上做基于时间点的database recover(时间点就是truncate时的时间点,可以用logmnr找出) 04:09:48 SQL> recover database until time '2011-10-27 04:01:50' using backup controlfile; ORA-00279: change 1354493 generated at 10/27/2011 04:00:23 needed for thread 1 ORA-00289: suggestion : /disk4/arch/test/arch_1_9_765501215.log ORA-00280: change 1354493 for thread 1 is in sequence #9 04:10:30 Specify log: {<RET>=suggested | filename | AUTO | CANCEL} auto Log applied. Media recovery complete. 12、open database 查看恢复 04:10:37 SQL> alter database open resetlogs; Database altered. 04:10:58 SQL> select * from scott.lxtb2; ID ---------- 1 2 3 4 5 6 7 8 10 11 10 rows selected. 04:11:30 SQL> ---------------------恢复成功 13、将恢复后的表空间导出(lxtbs2 tablespace) [oracle@work data]$ exp userid=\'sys/oracle as sydba\' point_in_time_recover=y tablespaces=lxtbs2 file=lxtbs2.dmp Export: Release 10.2.0.1.0 - Production on Thu Oct 27 04:15:34 2011 Copyright (c) 1982, 2005, Oracle. All rights reserved. EXP-00004: invalid username or password Username: sys as sysdba Password: Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production With the Partitioning, OLAP and Data Mining options Export done in US7ASCII character set and AL16UTF16 NCHAR character set server uses ZHS16GBK character set (possible charset conversion) Note: table data (rows) will not be exported About to export Tablespace Point-in-time Recovery objects... For tablespace LXTBS2 ... . exporting cluster definitions . exporting table definitions . . exporting table LXTB2 EXP-00091: Exporting questionable statistics. . exporting referential integrity constraints . exporting triggers . end point-in-time recovery Export terminated successfully with warnings. [oracle@work data]$ 14、将表空间导入到主库 ----------将主库表空间脱机 test (主库) 04:13:45 SQL> alter tablespace lxtbs2 offline; Tablespace altered. --------拷贝备库表空间datafile 到主库目录下 [oracle@work data]$ cp /u01/app/oracle/oradata/test2/lxtbs2.dbf /u01/app/oracle/oradata/test [oracle@work data]$ export ORACLE_SID=test -------------导入到主库 [oracle@work data]$ imp userid=\'sys/oracle as sysdba\' point_in_time_recover=y file=lxtbs2.dmp ignore=y Import: Release 10.2.0.1.0 - Production on Thu Oct 27 04:38:29 2011 Copyright (c) 1982, 2005, Oracle. All rights reserved. Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production With the Partitioning, OLAP and Data Mining options Export file created by EXPORT:V10.02.01 via conventional path About to import Tablespace Point-in-time Recovery objects... import done in US7ASCII character set and AL16UTF16 NCHAR character set import server uses ZHS16GBK character set (possible charset conversion) . importing SYS's objects into SYS . importing SCOTT's objects into SCOTT . . importing table "LXTB2" . importing SYS's objects into SYS Import terminated successfully without warnings 15、验证 04:30:39 SQL> alter tablespace lxtbs2 online; Tablespace altered. 04:38:51 SQL> select * from scott.lxtb2; ID ---------- 1 2 3 4 5 6 7 8 10 11 10 rows selected. -----------数据文件恢复到truncate 之前