ORA-00245 错误解决

从oracle 11.2.0.2.0开始，控制文件的快照文件必须能够被所有节点的数据实例访问到，并且必须放在共享的位置。否则，使用RMAN进行

今天看A类和B类的數據庫的日誌裏面都出現了如下錯誤：

節點1日誌：

Tue Jul 31 03:02:50 2012

Errors in file/soft/Oracle/diag/rdbms/a1qzdb/a1qzdb1/trace/a1qzdb1_ora_5374100.trc:

ORA-00245: control file backup operation failed

Tue Jul 31 09:08:37 2012

Thread 1 advanced to log sequence 346 (LGWRswitch)

Current log# 1 seq# 346 mem# 0:+DATA/a1qzdb/onlinelog/group_1.259.786961989

Tue Jul 31 09:08:37 2012

LNS: Standby redo logfile selected forthread 1 sequence 346 for destination LOG_ARCHIVE_DEST_2

Tue Jul 31 09:08:40 2012

Archived Log entry 1132 added for thread 1sequence 345 ID 0x1f7e84e dest 1:

Tue Jul 31 09:27:34 2012

Control file backup creation failed.

Backup target file size found to be zero.

Errors in file/soft/oracle/diag/rdbms/a1qzdb/a1qzdb1/trace/a1qzdb1_arc2_15073292.trc:

ORA-27037: unable to obtain file status

IBM AIX RISC System/6000 Error: 2: No such file or directory

Additional information: 3

節點2日誌：

Tue Jul 31 03:02:50 2012

Control file backup creation failed.

Backup target file size found to be zero.

Errors in file/soft/oracle/diag/rdbms/a1qzdb/a1qzdb2/trace/a1qzdb2_ckpt_7602620.trc:

ORA-27037: unable to obtain file status

IBM AIX RISC System/6000 Error: 2: No suchfile or directory

Additional information: 3

Tue Jul 31 09:16:27 2012

Thread 2 advanced to log sequence 284 (LGWRswitch)

Current log# 6 seq# 284 mem# 0:+DATA/a1qzdb/onlinelog/group_6.275.786964219

Tue Jul 31 09:16:27 2012

LNS: Standby redo logfile selected forthread 2 sequence 284 for destination LOG_ARCHIVE_DEST_2

Tue Jul 31 09:16:30 2012

Archived Log entry 1134 added for thread 2sequence 283 ID 0x1f7e84e dest 1:

Tue Jul 31 09:27:29 2012

Expanded controlfile section 19 from 818 to1636 records

Requested to grow by 818 records; added 2blocks of records

Tue Jul 31 09:27:34 2012

Errors in file/soft/oracle/diag/rdbms/a1qzdb/a1qzdb2/trace/a1qzdb2_ora_15204484.trc:

ORA-00245: control file backup operation failed

相應的trace日誌：

/soft/oracle/diag/rdbms/a1qzdb/a1qzdb1/trace/a1qzdb1_ora_5374100.trc:

*** 2012-07-31 03:02:50.674

*** SESSION ID:(541.7217) 2012-07-3103:02:50.675

*** CLIENT ID:() 2012-07-31 03:02:50.675

*** SERVICE NAME:(SYS$USERS) 2012-07-3103:02:50.675

*** MODULE NAME:(rman@qza1db1 (TNS V1-V3))2012-07-31 03:02:50.675

*** ACTION NAME:(0000011 STARTED114)2012-07-31 03:02:50.675

/soft/oracle/diag/rdbms/a1qzdb/a1qzdb1/trace/a1qzdb1_arc2_15073292.trc:

ORA-00245: control file backup operationfailed

*** 2012-07-31 09:27:34.107

IBM AIX RISC System/6000 Error: 2: No suchfile or directory

Additional information: 3

/soft/oracle/diag/rdbms/a1qzdb/a1qzdb2/trace/a1qzdb2_ckpt_7602620.trc:

*** 2012-07-31 09:27:34.382

*** SESSION ID:(885.58917) 2012-07-3109:27:34.383

*** CLIENT ID:() 2012-07-31 09:27:34.383

*** SERVICE NAME:(SYS$USERS) 2012-07-3109:27:34.383

*** MODULE NAME:(rman@qza1db2 (TNS V1-V3))2012-07-31 09:27:34.383

*** ACTION NAME:(0000021 STARTED114)2012-07-31 09:27:34.383

ORA-00245: control file backup operationfailed

/soft/oracle/diag/rdbms/a1qzdb/a1qzdb2/trace/a1qzdb2_ora_15204484.trc:

*** 2012-07-31 03:02:50.508

IBM AIX RISC System/6000 Error: 2: No suchfile or directory

Additional information: 3

Google了一下，發現這是Oracle11.2.0.2.0以後的一個“特性”：

从oracle 11.2.0.2.0开始，控制文件的快照文件必须能够被所有节点的数据实例访问到，并且必须放在共享的位置。否则，使用RMAN进行备份控制文件的时候会报如下错误：

ORA-00245: control file backup operationfailed。

当使用备份脚本备份的时候，，这个错误可能会导致备份工作中断。

METALINK 中的两篇文档：

RAC BACKUP FAILS WITH ORA-00245: CONTROLFILE BACKUP OPERATION FAILED [ID1268725.1]【這篇文字沒找到】

RMAN Snapshot Controlfile Must Reside onShared Device for RAC database in 11G [ID 1263621.1]【這篇文字沒找到】

解释并解决了此问题！摘要如下：

"RMAN creates a copy of the controlfile for read consistency, this is the snapshot controlfile. Due to the changesmade to the controlfile backup mechanism in 11gR2 any instances in the clustermay write to the snapshot controlfile. Therefore, the snapshot controlfile fileneeds to be visible to all instances.

The same happens when a backup of thecontrolfile is created directly from sqlplus any instance in the cluster maywrite to the backup controfile file.

In 11gR2 onwards, the controlfile backuphappens without holding the control file enqueue. For non-RAC database, thisdoesn't change anything.

But, for RAC database, the snapshotcontrolfile location must be in a shared file system that will be accessible fromall the nodes."

当前的控制文件配置：

RMAN> show all;

using target database control file insteadof recovery catalog

RMAN configuration parameters for databasewith db_unique_name OSDB are:

CONFIGURE RETENTION POLICY TO REDUNDANCY 1;# default

CONFIGURE BACKUP OPTIMIZATION OFF; #default

CONFIGURE DEFAULT DEVICE TYPE TO DISK; #default

CONFIGURE CONTROLFILE AUTOBACKUP OFF; #default

CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FORDEVICE TYPE DISK TO '%F'; # default

CONFIGURE DEVICE TYPE DISK PARALLELISM 1BACKUP TYPE TO BACKUPSET; # default

CONFIGURE DATAFILE BACKUP COPIES FOR DEVICETYPE DISK TO 1; # default

CONFIGURE ARCHIVELOG BACKUP COPIES FORDEVICE TYPE DISK TO 1; # default

CONFIGURE MAXSETSIZE TO UNLIMITED; #default

CONFIGURE ENCRYPTION FOR DATABASE OFF; #default

CONFIGURE ENCRYPTION ALGORITHM 'AES128'; #default

CONFIGURE COMPRESSION ALGORITHM 'BASIC' ASOF RELEASE 'DEFAULT' OPTIMIZE FOR LOAD TRUE ; # default

CONFIGURE ARCHIVELOG DELETION POLICY TONONE; # default

CONFIGURE SNAPSHOT CONTROLFILE NAME TO'/soft/product/11.2.0.2/dbh/dbs/snapcf_osdb1.f'; # default

由於我們使用的是ASM，所以可以使用如下配置解决此问题，也可以把ASM替換成共享目錄：

RMAN> CONFIGURE SNAPSHOT CONTROLFILE NAME TO'+SFRA/sncforcl_osdb.ora';

new RMAN configuration parameters:

CONFIGURE SNAPSHOT CONTROLFILE NAME TO '+SFRA/sncforcl_osdb.ora';

new RMAN configuration parameters aresuccessfully stored

RMAN> show all;

RMAN configuration parameters for databasewith db_unique_name OSDB are:

CONFIGURE RETENTION POLICY TO REDUNDANCY 1;# default

CONFIGURE BACKUP OPTIMIZATION OFF; #default

CONFIGURE DEFAULT DEVICE TYPE TO DISK; #default

CONFIGURE CONTROLFILE AUTOBACKUP OFF; #default

CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FORDEVICE TYPE DISK TO '%F'; # default

CONFIGURE DEVICE TYPE DISK PARALLELISM 1BACKUP TYPE TO BACKUPSET; # default

CONFIGURE DATAFILE BACKUP COPIES FOR DEVICETYPE DISK TO 1; # default

CONFIGURE ARCHIVELOG BACKUP COPIES FORDEVICE TYPE DISK TO 1; # default

CONFIGURE MAXSETSIZE TO UNLIMITED; #default

CONFIGURE ENCRYPTION FOR DATABASE OFF; #default

CONFIGURE ENCRYPTION ALGORITHM 'AES128'; #default

CONFIGURE COMPRESSION ALGORITHM 'BASIC' ASOF RELEASE 'DEFAULT' OPTIMIZE FOR LOAD TRUE ; # default

CONFIGURE ARCHIVELOG DELETION POLICY TONONE; # default

CONFIGURE SNAPSHOT CONTROLFILE NAME TO '+SFRA/sncforcl_osdb.ora';

附件：

In RAC environment from 11.2 onwards Backup Or Snapshot controlfile needs to be in shared location [ID 1472171.1]