How to Migrate Oracle 10.2 32bit to 10.2 64bit on Microsoft Windows

Action Plan for Migrate.

1 Take a cold backup and export full db or important schema
2 Take a backup of 10g binaries (Oracle Home)
3 Take a control file backup using command
Alter database backup controlfile to trace;
4 Take a backup of init.ora file
5 Prepare the current oracle directory structure
6 Clone Prod to Prod (32 bit O/S)
7 Install 64 bit Win O/S
8 Create existing oracle directory structure
9 Install 10g (10.2.0.1 Base) oracle binaries(Software Only),Upgrade 10.2.0.4 If you use exp/imp install with database
10 Please follow the Scenario 1 or 2 or 3



Scenario 1:


Oracle 10.2.0.1(32 Bit) to Oracle 10.2.0.1(64 Bit)

1. Install Oracle 10.2.0.1.0 64bit, on Windows 2003 64bit (either AMD64 or EM64T processors).
2. The same directory structure was used for both machines.
NOTE: If a different directory structure will be used, then you will need to recreate the controlfiles, before starting the migration process.

Alter database backup controlfiles to trace;

3. Copy the Database files from the 32bit machine, to the 64bit machine.
(Datafiles, Logfiles, Controlfiles, init.ora, tnsnames.ora, listener.ora, etc...)

4. Create the Oracle service on the 64bit system, using the ORADIM command.

(Eg, oradim –new –sid orcl –startmode auto –pfile

6.Run Sqlplus

sqlplus '/as sysdba'

SQL> startup upgrade

SQL>@$Oracle_Home\rdbms\admin\utlip.sql

SQL>@$Oracle_Home\rdbms\admin\utlrp.sql

SQL> shut immediate

SQL>startup

SQL>select name,platform_name,open_mode from v$database;


Scenario 2:


Oracle 10.2.0.1(32 Bit) to Oracle 10.2.0.4(64 Bit)


1. Install Oracle 10.2.0.1.0 64bit, on Windows 2003 64bit (either AMD64 or EM64T processors).

2. If you wish to apply a patchset, you can do that now. For this test, the 10.2.0.4 patchset was applied to the 64bit installation.

3. The same directory structure was used for both machines.
NOTE: If a different directory structure will be used, then you will need to recreate the controlfiles, before starting the migration process.

Alter database backup controlfile to trace;

4. Copy the Database files from the 32bit machine, to the 64bit machine.
(Datafiles, Logfiles, Controlfiles, init.ora, tnsnames.ora, listener.ora, etc..)

5. Create the Oracle service on the 64bit system, using the ORADIM command.
(Eg, oradim –new –sid orcl –startmode auto –pfile
6.Run Sqlplus

sqlplus '/as sysdba'

SQL> startup upgrade

SQL>@$Oracle_Home\rdbms\admin\utlip.sql

SQL>@$Oracle_Home\rdbms\admin\catupgrd.sql

SQL>@$Oracle_Home\rdbms\admin\utlrp.sql

SQL> shut immediate

SQL>startup

SQL>select name,platform_name,open_mode from v$database;


Note: 403522.1:- How to Migrate Oracle 10.2 32bit to 10.2 64bit on Microsoft Windows


Scenario 3:

Steps to migrate 32 bit oracle 9i database to 64 bit 10g database:

1, Make backups of 32 bit 10g database.
Cold backup, export.

You need to have export file for migration. You can use user system to export entire database.

2, Install 64 bit Oracle 10g.

3, Create required tablespace, users etc.

4, Import 10g export dump file to 10g (from user=user1 to user=user1).


Steps

CONN sys/password@SID_NAME AS SYSDBA

1. Create Schema Name

2. ALTER USER U_NAME IDENTIFIED BY PASSWD ACCOUNT UNLOCK;
GRANT CREATE ANY DIRECTORY TO ;

3. CREATE OR REPLACE DIRECTORY DIR_NAME AS '/u05/oradata/';

4. GRANT READ, WRITE ON DIRECTORY DIR_NAME TO SCHEMA_NAME;

step:2

Schema Exports/Imports

expdp USER_NAME/PASSWD@SID_NAME schemas=SCHEMA_NAME directory=DIR_NAME dumpfile=INV.dmp logfile=EXPINV.log

impdp USER_NAME/PASSWD@SID_NAME schemas=SCHEMA_NAME directory=DIR_NAME dumpfile=INV.dmp logfile=IMPINV.log


Full Database Exports/Imports

expdp sys/password@SID_NAME full=Y directory=DIR_NAME dumpfile=FULLDB.dmp logfile=FULLDB10G.log

impdp sys/password@SID_NAME full=Y directory=DIR_NAME dumpfile=FULLDB.dmp logfile=FULLDB10G.log

Enterprise Manager DST Problem

Purpose

OEM agent suddently not running for daylight saving time changes.


Problem:

$ ./emctl start agent

Oracle Enterprise Manager 10g Release 10.2.0.1.0.

Copyright (c) 1996, 2005 Oracle Corporation. All rights reserved.

Starting agent ..... failed.

The agentTZRegion value in /oracle/app/em/agent10g/sysman/config/emd.properties is not in agreement with what agent thinks it should be.Please verify your environment to make sure that TZ setting has not changed since the last start of the agent.

If you modified the timezone setting in the environment, please stop the agent and exectute 'emctl resetTZ agent' and also execute the script mgmt_target.set_agent_tzrgn(, ) to get the value propagated to repository.

Consult the log files in: /oracle/app/em/agent10g/sysman/log


Solutions:

Download the Patch 5632264 for the latest version for your platform, DO NOT APPLY IT with opatch but:

Manually replace the oracore [/zoneinfo/timez*.dat] files in the Central Agent ORACLE_HOME:
Agent ORACLE_HOME

- backup the initial oracore files $ORACLE_HOME/oracore/[/zoneinfo/timez*.dat]

- replace the oracore [/zoneinfo/timez*.dat] files in the Central Agent ORACLE_HOME with the oracore [/zoneinfo/timez*.dat] files extracted from the patch


- ./emctl resetTZ agent

Oracle Enterprise Manager 10g Release 10.2.0.1.0.
Copyright (c) 1996, 2005 Oracle Corporation. All rights reserved.
Updating /oracle/oracle/agent10g/sysman/config/emd.properties...
Successfully updated /oracle/oracle/agent10g/sysman/config/emd.properties.
Login as the em repository user and run the script:
exec mgmt_target.set_agent_tzrgn('oradiscprod.amestruetemper.com:3872','US/Eastern')
This can be done for example by logging into sqlplus and doing
SQL> exec mgmt_target.set_agent_tzrgn('oradiscprod.amestruetemper.com:3872','US/Eastern')

- start the Central Agent

./emctl start agent
./emctl status agent


Reference:
Enterprise Manager DST Quick Fix Guide: 418792.1

index is in unusable state

Error Messsage

While compiling the flexfields, the procedure wf_event.raise failed with following
parameters:
event_name = oracle.apps.fnd.flex.dff.compiled
event_key = FND.$SRS$.FNDSCURS

The error raised is:
ORA-00604: error occurred at recursive SQL level 1
ORA-01502: index 'APPLSYS.SYS_C00169351' or partition of such index is in unusable state

Check the Unstale index

select index_name, status from dba_indexes where tablespace_name='APPS_TS_QUEUES' and STATUS != 'VALID';

WF_JAVA_DEFERRED_N1 UNUSABLE
SYS_C00169351 UNUSABLE
WF_DEFERRED_N1 UNUSABLE
SYS_C00105999 UNUSABLE


Solution:

Rebuild Unstale index.


select 'alter index 'owner'.'index_name' rebuild;' from dba_indexes where tablespace_name='APPS_TS_QUEUES' and STATUS != 'VALID';

SQL*Plus Command Reference

SQL*PLUS - SET Statement

http://www.ss64.com/orasyntax/plus_set.html

SQL*Plus FAQ
http://www.orafaq.com/wiki/SQL*Plus_FAQ

SQL*Plus Command Reference
http://download-west.oracle.com/docs/cd/B10500_01/server.920/a90842/ch13.htm

How to Restart The Concurrent Manager in Unix.

goal: How to restart the concurrent manager in Unix
goal: How to bounce the concurrent manager in Unix
fact: Oracle Application Object Library


fix:

1. At the command line issue the following shutdown command:


CONCSUB apps/apps sysadmin 'System Administrator' SYSADMIN CONCURRENT FND SHUTDOWN


Note: There are 3 options which can be used for controlling the Internal
Manager from the OS as follows:


SHUTDOWN = Standard, non-invasive shutdown.

DEACTIVATE = Will not allow further Requests to Process once Issued.

ABORT = Terminate Requests and Deactivate the Internal Manager.



2. At the command line, issue the following Start Up command:

startmgr sysmgr="[fnd_usernamd/fnd_password]" mgrname="[mgrname]" printer="[printername]"
mailto="[userid userid2...]" restart="[Nminutes]" logfile="[log_file_name]" sleep="
[new_check]" pmon="[manager_check]" quesiz="[number_check]" diag="[YN]"


ExAMPLE:

startmgr sysmgr="applsys/apps" mgrname="std" printer="localprint"
mailto="Ramesh Mani" restart="N" logfile="mgrlog" sleep="90"
pmon="5" quesiz="10" diag="N"



Note: a space is required between arguments and the "diag" parameter
may be omitted if desired.

Reference: 147449.1

The field PARAMETER.CONFIG could not be located or read

Symptoms

You are encountering the following error while trying to start/stop the apache
using adapcctl.sh


"The field PARAMETER.CONFIG could not be located or read."

The adapcctl.log carries the following errors:

Cannot read value from field PARAMETER.CONFIG

Cause: The field PARAMETER.CONFIG could not be located or read.


Changes

You upgraded to RDBMS 9.2.0.4 from RDBMS 8.0.6, but the instance was fine for a couple of days after the upgrade. This problem suddenly cropped up.

Cause

Your guest user is end-dated
That is the sole reason you are getting this error.

Fix

Back up your FND_USER TABLE

Run the following command update fnd_user set end_date=null where user_name='GUEST';

Retest with adapcctl.sh


Reference: 298143.1