又见Oracle BUG:Rman备份的alert错误提示

又见Oracle BUG:Rman备份的alert错误提示

又见Oracle BUG:Rman备份的alert错误提示
 现象：
 在执行数据库备份时，alert日志提示存在错误信息，并给出了一个详细错误信息的trc文件。

推荐阅读：

RMAN 配置归档日志删除策略

Oracle基础教程之通过RMAN复制数据库

RMAN备份策略制定参考内容

RMAN备份学习笔记

Oracle数据库备份加密 RMAN加密
 
alert.log
 <<====
 Tue May 7 10:23:19 2013
 Errors in file /opt/oracle/admin/MMS/udump/MMS_ora_9140.trc:
 Tue May 7 10:23:19 2013
 Errors in file /opt/oracle/admin/MMS/udump/MMS_ora_9140.trc:
 Tue May 7 10:23:19 2013
 Errors in file /opt/oracle/admin/MMS/udump/MMS_ora_9140.trc:
 <<====
 
但trc文件没有异常信息：
 [oracle@MMSDB admin]$ more /opt/oracle/admin/MMS/udump/MMS_ora_9140.trc
 /opt/oracle/admin/MMS/udump/MMS_ora_9140.trc
 Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bit Production
 With the Partitioning, OLAP and Data Mining options
 ORACLE_HOME = /opt/oracle/product/10.2.0/db_1
 System name: Linux
 Node name: MMSDBZ
 Release: 2.6.18-194.el5
 Version: #1 SMP Tue Mar 16 21:52:39 EDT 2010
 Machine: x86_64
 Instance name: MMS
 Redo thread mounted by this instance: 1
 Oracle process number: 86
 Unix process pid: 9140, image: oracleMMS@MMSDBZ

*** 2013-05-07 10:23:19.906
 *** ACTION NAME:(0000089 STARTED16) 2013-05-07 10:23:19.873
 *** MODULE NAME:(backup full datafile) 2013-05-07 10:23:19.873
 *** SERVICE NAME:(MMS) 2013-05-07 10:23:19.873
 *** SESSION ID:(891.56220) 2013-05-07 10:23:19.873
 *** ACTION NAME:(0000093 STARTED111) 2013-05-07 10:23:24.597
 [oracle@MMSDB admin]$
 
分析：
 1.确认alert错误提示发生在执行备份时；
 MMS_ora_9140.trc：
 *** MODULE NAME:(backup full datafile) 2013-05-07 10:23:19.873
 2.trc文件是否被截断？
 3.是否是Oracle本身存在BUG。
 
对于1，查看了备份脚本和rman日志，确认备份脚本设置正确，无异常rman日志。排除1
 对于2，查看了max_dump_file_size参数设置：
 SQL> show parameter max_dump_file_size
 NAME TYPE VALUE
 ------------------------------------ -------------------------------- ------------------------------
 max_dump_file_size string UNLIMITED  <===设置没问题
 查看操作系统空间也是有的。
 对于3，在Metalink查找，发现有BUG存在。
 Trace File Created when RMAN issues the Controlfile Autobackup [ID 335731.1]
 
Applied to:
 Oracle Database - Enterprise Edition - Version 10.1.0.2 to 10.2.0.1 [Release 10.1 to 10.2]  <====我们的DB版本为10.2.0.1
 Information in this document applies to any platform.
 ***Checked for relevance on 03-Apr-2013***
 
 
 
Symptoms:
 Error in Alert.log on controlfile autobackup and trace file created with RMAN backups.
 
Issue verified by alert.log showing error:
 Errors in file /usr/oracle/product/admin/dbtest/udump/dbtest_ora_8477.trc:
 Control autobackup written to DISK device
 handle '/orabackup/backup/prod/c-959859059-20050906-01'
 
And trace file created with:
 
/usr/oracle/product/admin/dbtest/udump/dbtest_ora_8477.trc
 Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
 With the Partitioning, OLAP and Data Mining options
 ORACLE_HOME = /usr/oracle/product/10.2
 System name: Linux
 Node name: dbtest
 Release: 2.6.9-11.ELsmp
 Version: #1 SMP Fri May 20 18:26:27 EDT 2005
 Machine: i686
 Instance name: dbtest
 Redo thread mounted by this instance: 1
 Oracle process number: 26
 Unix process pid: 8477, image: oracle@dbtest (TNS V1-V3)
 
*** 2005-09-06 11:22:02.425
 *** ACTION NAME:(0000064 STARTED111) 2005-09-06 11:22:02.403
 *** MODULE NAME:(backup full datafile) 2005-09-06 11:22:02.403
 *** SERVICE NAME:(SYS$USERS) 2005-09-06 11:22:02.403
 *** SESSION ID:(538.56240) 2005-09-06 11:22:02.403
 *** 2005-09-06 11:27:00.173
 *** ACTION NAME:(0000670 STARTED111) 2005-09-06 11:27:00.173
 *** MODULE NAME:(backup archivelog) 2005-09-06 11:27:00.173
 Cause:
 Bug 4596065.  Development determined these empty trace files are created upon controlfile autobackup when flash recovery area is enabled.
 Solution:
 To implement the solution, please execute the following steps:
 1. Ignore the trace files until patch can be applied.
 
10.2.0.2 Patch Set - List of Bug Fixes by Problem Type [ID 358749.1]  <===没有one-off patch,该BUG在10.0.2.2中被修复
 
解决方法：
 忽略该错误提示，或者升级数据库到10.2.0.2及以上。