ORA-06512: at "ORACLE_OCM.MGMT_DB_LL_METRICS"

来源：互联网发布：房地产网站源码编辑：程序博客网时间：2024/05/07 13:53

一、2014/7/8早上检查数据库的alert日志发现如下报错：

Mon Jul 07 22:00:04 2014
Errors in file /apps/oracle/diag/rdbms/egapdb/egapdb1/trace/egapdb1_j001_28246242.trc:
ORA-12012: error on auto execute of job "ORACLE_OCM"."MGMT_CONFIG_JOB_2_1"
ORA-29280: invalid directory path
ORA-06512: at "ORACLE_OCM.MGMT_DB_LL_METRICS", line 2436
ORA-06512: at line 1
Mon Jul 07 22:05:11 2014

trace文件中内容如下：

vi /apps/oracle/diag/rdbms/egapdb/egapdb1/trace/egapdb1_j001_28246242.trc
"/apps/oracle/diag/rdbms/egapdb/egapdb1/trace/egapdb1_j001_28246242.trc" 27 lines, 1023 characters
Trace file /apps/oracle/diag/rdbms/egapdb/egapdb1/trace/egapdb1_j001_28246242.trc
Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production
With the Partitioning, Real Application Clusters, OLAP, Data Mining
and Real Application Testing options
ORACLE_HOME = /apps/oracle/product/11.2.0.3/db_1
System name:    AIX
Node name:      egapdb1
Release:        1
Version:        6
Machine:        00F8284F4C00
Instance name: egapdb1
Redo thread mounted by this instance: 1
Oracle process number: 869
Unix process pid: 28246242, image: oracle@egapdb1 (J001)

*** 2014-07-07 22:00:04.620
*** SESSION ID:(642.65245) 2014-07-07 22:00:04.620
*** CLIENT ID:() 2014-07-07 22:00:04.620
*** SERVICE NAME:(SYS$USERS) 2014-07-07 22:00:04.620
*** MODULE NAME:(DBMS_SCHEDULER) 2014-07-07 22:00:04.620
*** ACTION NAME:(MGMT_CONFIG_JOB_2_1) 2014-07-07 22:00:04.620

ORA-12012: error on auto execute of job "ORACLE_OCM"."MGMT_CONFIG_JOB_2_1"
ORA-29280: invalid directory path
ORA-06512: at "ORACLE_OCM.MGMT_DB_LL_METRICS", line 2436
ORA-06512: at line 1

二、此套RAC的OS版本及DB版本狀況：

SQL> select * from v$version; （打了psu5补丁）

BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production
PL/SQL Release 11.2.0.3.0 - Production
CORE 11.2.0.3.0 Production
TNS for IBM/AIX RISC System/6000: Version 11.2.0.3.0 - Production
NLSRTL Version 11.2.0.3.0 - Production

操作系统版本：

uname -a
AIX egapdb1 1 6 00F8284F4C00

三、网上有很多该问题的解决方案。

http://www.linuxidc.com/Linux/2013-01/78399.htm

http://blog.csdn.net/tianlesoftware/article/details/8532090

內容摘錄如下：

Alert log 出现如下错误：

Errors in file /Oracle/11.2.0.3/oracle/diag/rdbms/fssms/fssms2/trace/fssms2_j002_17105234.trc:
ORA-12012: error on auto execute of job "ORACLE_OCM"."MGMT_CONFIG_JOB_2_2"
ORA-29280: invalid directory path
ORA-06512: at "ORACLE_OCM.MGMT_DB_LL_METRICS", line 2436
ORA-06512: at line 1

操作系统： AIX

数据库：11.2.0.3, RAC

这个报错的是ORACLE_OCM的这个用户，这个用户是Oracle 预定义的非管理员用户。这个在我之前的文章里有说明：

Oracle 用户及角色介绍 http://www.linuxidc.com/Linux/2013-01/78398.htm

ORACLE_OCM
The account used with Oracle Configuration Manager. This feature enables you to associate the configuration information for the current Oracle Database instance with OracleMetaLink. Then when you log a service request, it is associated with the database instance configuration information.

See Oracle Database Installation Guide for your platform.
Expired and locked

ORACLE_OCM 账户是用来管理Oracle Configuration Manager的。默认情况下，该账户是Expired and locked状态。

该错误只出现在OracleRDBMS 11.2.0.3 的版本（新安装的db，或者从其他版本升级到11.2.0.3的），或者Oracle Configuration Manager 的版本大于10.3.3.0.0(opatch lsinventory -detail 可以查看其版本)。

导致这个错误的原因如下：

Newer databasereleases are automatically instrumented for OCM collections. Buthere, the OCM instrumentation job is trying to write to a "state"directory which doesn't yet exists.

当自动执行的job发现目录不存在后，就会报错。

可以用sysdba用户执行如下SQL，来验证OCM db 目录：

SQL> set lin 160

SQL> col owner for a10

SQL> col DIRECTORY_NAME for a25

SQL> col DIRECTORY_PATH for a50

SQL> select * from dba_directories where DIRECTORY_NAME like '%OCM_CONFIG%';

OWNER DIRECTORY_NAME DIRECTORY_PATH

---------- ---------------------------------------------------------------------------

SYS ORACLE_OCM_CONFIG_DIR /u01/app/oracle/product/11.2.0/dbhome_1/ccr/state

这里只有一个ORACLE_OCM_CONFIG_DIR的目录，而MGMT_DB_LL_METRICS需要写入目录：ORACLE_OCM_CONFIG_DIR2。但是这个目录在built-in 的instrumentation脚本中并没有创建，所以才会包我们的这个错误。

解决方法1：

执行"ORACLE_HOME/ccr/admin/scripts/installCCRSQL"脚本重新配置OCM collections。

如果没有这个脚本，那么需要先运行"ORACLE_HOME/ccr/bin/setupCCR"脚本来配置OCM。

注意：如果是non-shared文件系统的RAC 环境，在运行installCCRSQL.sh脚本之前必须先指定变量ORACLE_CONFIG_HOME 到$ORACLE_HOME。具体可以参考：

OracleConfiguration Manager Installation and Administration Guide

http://docs.oracle.com/cd/E28601_01/doc.1037/e26167/install.htm#CBHFFGFJ

解决方法2：

如果不使用OCM Collector, 可以从数据库中移除OCM的对象。

SQL> drop user ORACLE_OCM cascade;

OCM 是一个独立的工具，移除OCM 不影响数据库其他功能的使用。

如果不删除OCM 的对象，也可以禁用OCM的jobs，命令如下：

execdbms_scheduler.disable('ORACLE_OCM.MGMT_CONFIG_JOB')
exec dbms_scheduler.disable('ORACLE_OCM.MGMT_STATS_CONFIG_JOB')

MOS的文档，把该错误定位成了Bug 13385346：

Bug 13385346 : GCR0 TRACE FILES CONTINUOUSLY CREATED WITHKJGCR_DELETESO MESSAGES

该bug的解决方式是：忽略生成的错误信息。该bug在12.1中已经修复。

MOS 参考文档：

"ORA-12012: error on auto execute ofjob ORACLE_OCM.MGMT_CONFIG_JOB_2_1" And "ORA-29280: invalid directorypath" In Database AlertLog [ID 1453959.1]

How To Manually Remove ORACLE_OCM FromDatabase [ID 859113.1]

四：MOS上与该问题相关NOTE

Bug 13385346 : GCR0 TRACE FILES CONTINUOUSLY CREATED WITHKJGCR_DELETESO MESSAGES

"ORA-12012: error on auto execute ofjob ORACLE_OCM.MGMT_CONFIG_JOB_2_1" And "ORA-29280: invalid directorypath" In Database AlertLog [ID 1453959.1]

How To Manually Remove ORACLE_OCM FromDatabase [ID 859113.1]

五：建议按照MOS文档介绍的方案解决ORACLE相关问题

0 0