怎么恢复没有备份的Oracle数据库

海外服务器 (627) 2015-11-11 14:21:25

数据文件丢失,没有备份,拥有文件创建以来的全部归档,使用RMAN恢复,报错RMAN-06102: no channel to restore a backup or copy of log thread 1 seq 726 scn 1757142927; 使用sqlplus恢复, 执行 'Alter Database recover datafile ' Fails with ORA-279。

 

总结: RMAN备份没有使用catalog, controlfile默认保留7天的备份/归档信息,v$archived_log没有记录足够多的归档信息, 所以报RMAN-06102, 需要通过CATALOG命令注册。

 

SQLPLUS 执行 'Alter Database recover datafile ' Fails with ORA-279 因为9.2.0.6版本使用这个命令不能自动执行recover. 用户版本9.2.0.1也遇到同样的问题,使用RECOVER DATAFILE即可。

 

处理步骤:

1、生成controlfile 备份到文本ora9roro_ora_479268.trc

2、查询该文件创建时间为Feb 21 11:10:24 2006 (查询alert.log)

Tue Feb 21 11:10:24 2006  

CREATE TABLESPACE tzgl DATAFILE 'tzgl' SIZE 100M  

Tue Feb 21 11:10:29 2006  

Completed: CREATE TABLESPACE tzgl DATAFILE 'tzgl' SIZE 100M  

Tue Feb 21 21:58:23 2006  

Thread 1 advanced to log sequence 277  

Current log# 3 seq# 277 mem# 0: /oradata/ora9roro/redo03.log  

Tue Feb 21 21:58:23 2006 

3. 用户说没有备份(没有确实检查是否有备份),创建空文件,因为有全部的归档文件存在,考虑建空文件, 使用归档文件恢复.

alter database create datafile '/oracle/product/9.2.0/dbs/tzgl' as '/oracle/product/9.2.0/dbs/tzgl'; 

或者就写alter database create datafile 12 as '/oracle/product/9.2.0/dbs/tzgl';

该文件file#=12.

4.rman target /

rman> recover datafile 12; 失败  

archive log thread 1 sequence 1116 is already on disk as file /oradata/ora9roro/archive/1_1116.dbf  

archive log thread 1 sequence 1117 is already on disk as file /oradata/ora9roro/archive/1_1117.dbf  

archive log thread 1 sequence 1118 is already on disk as file /oradata/ora9roro/archive/1_1118.dbf  

....  

RMAN-06102: no channel to restore a backup or copy of log thread 1 seq 728 scn 1757357012  

RMAN-06102: no channel to restore a backup or copy of log thread 1 seq 727 scn 1757357009  

RMAN-06102: no channel to restore a backup or copy of log thread 1 seq 726 scn 1757142927 

检查归档文件是否存在, 发现从1_1.dbf直到现在, 所有的ARCHIVELOGS都存在.但RMAN-06102错误信息表明controlfile并没有记录下所有的归档信息.查看参数control_file_record_keep_time 为默认值7. 所以只保留7天的备份信息.

调整control_file_record_keep_time=365以保证以后的备份可以保留更长的时间.

企图注册归档文件到controlfile. 但不支持, 只针对standby controlfile.

alter database register logfile '/oradata/ora9roro/archive/1_726.dbf';  

ERROR at line 1:  

ORA-01665: controlfile is not a standby controlfile 

7. 尝试通过sqlplus 恢复, 也失败. ---但这里并没有查看原因, 想当然认为rman恢复不行,sqlplus 执行也不行. 在step 16找到了原因.

SQL> conn / as sysdba  

Connected.  

SQL> alter database recover datafile 12;  

alter database recover datafile 12  

*  

ERROR at line 1:  

ORA-00279: change 1181419363 generated at 02/21/2006 11:10:29 needed for thread  

1  

ORA-00289: suggestion : /oradata/ora9roro/archive/1_276.dbf  

ORA-00280: change 1181419363 for thread 1 is in sequence #276 

8. 验证问题是否出在归档文件, 尝试移动/oradata/ora9roro/archive/1_276.dbf到其他目录, 然后进行恢复alter database recover datafile 12;

错误同上.所以, 不是归档文件本身故障的问题. 继续查无法读取归档文件的原因.

9. 查看是否重建过controlfile或open resetlogs操作.

从2006年2月21日开始,有没有做过重建controlfile或open resetlogs之类的操作?

select RESETLOGS_TIME from v$database;  

RESETLOGS  

--------- 

-APR-05

结果表示, 没有resetlog过, 也没有重建过controlfile

10.查看v$bakcup_piece, v$backup_datafile发现有记录, 是2006年12月29日的备份记录,备份到带库.

原来并非像用户所说没有备份

root身份查看

#crontab -l找到备份,每天执行,可是查看/tmp/bkdb_$dt.log文件, 最后一次成功备份是bkdb_200612250100.log

0 1 * * * /usr/tivoli/tsm/client/oracle/orabackup.sh >/dev/null#  

# cat /usr/tivoli/tsm/client/oracle/orabackup.sh  

export dt=`date +%Y%m%d%H%M`  

su - oracle -c "rman target / nocatalog cmdfile /oracle/sched/bkdb.scr msglog /tmp/bkdb_$dt.log"  

/oracle/sched/bkdb.scr备份脚本内容: bkdb_200612290100.log  

RUN{  

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE' parms =  

'ENV=(TDPO_OPTFILE=/usr/tivoli/tsm/client/oracle/bin64/tdpo.opt)';  

BACKUP  

FORMAT 'df_T%T_s%s_p%p_t%t'  

FILESPERSET 2  

DATABASE;  

RELEASE CHANNEL ch00;  

11. 尝试恢复:

RMAN> RUN{  

ALLOCATE CHANNEL ch00 TYPE 'SBT_TAPE' parms =  
'EN2> 3> V=(TDPO_OPTFILE=/usr/tivoli/tsm/client/oracle/bin64/tdpo.opt)';  

4> restore datafile 12;  

5> release channel ch00  

6> ;  

7> }  

using target database controlfile instead of recovery catalog  

allocated channel: ch00  

channel ch00: sid=11 devtype=SBT_TAPE 

channel ch00: Tivoli Data Protection for Oracle: version 5.2.0.0   

Starting restore at 07-APR-08  

channel ch00: starting datafile backupset restore  

channel ch00: specifying datafile(s) to restore from backup set  

restoring datafile 00012 to /oracle/product/9.2.0/dbs/tzgl  

released channel: ch00  

RMAN-00571: ===========================================================  

RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============  

RMAN-00571: ===========================================================  

RMAN-03002: failure of restore command at 04/07/2008 14:20:09  

ORA-19501: read error on file "df_T20061229_s2660_p1_t610419667", blockno 1 (blocksize=512)  

ORA-27190: skgfrd: sbtread2 returned error  

ORA-19511: Error received from media manager layer, error text:  

ANS1314E (RC14) File data currently unavailable on server 

带库有问题, 或者数据已经被覆写. 所以, 不再考虑使用备份做恢复.

13. 查看v$recover_file, v$recovery_log, v$log_history,看系统状态,信息存储在recover.lst中

set pagesize 20000  

set linesize 180  

set pause off  

set serveroutput on  

set feedback on  

set echo on  

set numformat 999999999999999  

Spool recover.lst  

select substr(name, 1, 50), status from v$datafile;  

select file#,substr(name,1,50), recover, fuzzy, to_char(checkpoint_time,'dd/mm/yyyy:hh24:mi:ss') ckpt_time, checkpo  

int_change#, resetlogs_change#, to_char(resetlogs_time,'dd/mm/yyyy HH24:MI:SS')  

tm from v$datafile_header;  

select * from v$backup;  

select name, open_mode, checkpoint_change#, ARCHIVE_CHANGE# from v$database;  

select GROUP#,THREAD#,SEQUENCE#,MEMBERS,ARCHIVED,STATUS,FIRST_CHANGE# from v$log;  

select GROUP#,substr(member,1,60) from v$logfile;  

select * from v$log_history;  

select * from v$recover_file;  

select * from v$recovery_log;  

select HXFIL File_num,substr(HXFNM,1,40) File_name,FHTYP Type,HXERR Validity,  

FHSCN SCN, FHTNM TABLESPACE_NAME,FHSTA status ,FHRBA_SEQ Sequence   

from X$KCVFH;  

spool off 

 

14恢复方案:

 

alter database create datafile 12 as '/oracle/product/9.2.0/dbs/tzgl'; ----ÒѾ×öÁËin step3  

recover datafile 12;  

now start applying archives from the time of creation of datafile.  

once the archives are applied and redo logs are applied, you can issue the command, "alter database open". 

 

15. 注册归档文件到controlfile (和catalog db 可选)
 

rman target /  

RMAN> list archivelog all; --[1_1116.dbf, 1_1164.dbf]  

RMAN> CATALOG ARCHIVELOG '/oradata/ora9roro/archive/1_1.dbf'; ---添加到v$archived_log  

RMAN> list archivelog all; --多显示1_1.dbf记录  

Key Thrd Seq S Low Time Name  
------- ---- ------- - --------- ----  
1165 1 1 A 16-APR-05 /oradata/ora9roro/archive/1_1.dbf 

 

16. 在step7中登陆sqlplus恢复失败, 查原因:

metalink<352617.1> 'Alter Database recover datafile ' Fails with ORA-279  

Noticable difference between 9.2.0.6 and other versions is that the "recover datafile" syntax does not auotmatically   

perform auto recover, Oracle prompts for each log. On other versions this is not the case.  

This is reported as Bug: 4178579 - ALTER DATABASE RECOVER DATAFILE IS BROKEN IN 9.2.0.6 

 

解决方法:

Use different syntax:  

recover datafile x   

recover automatic datafile x  

alter database recover automatic datafile x  

或者打补丁:Apply 9.2.0.7.0 patchset 

THE END