Date: 4th Aug 2010

HOTFIX: DBED SUPPORT FOR ORACLE 11G R2

OS: AIX
OS Version: 5.3, 6.1
Etrack Incidents:2068232

Errors/Problems Fixed: Added support for Oracle 11g R2 to DBED utilities (2068232)

OVERVIEW
----------------
This hotfix adds support for Oracle 11g R2 on Veritas Storage Foundation for Oracle 5.0 MP3RP4 on AIX.
 
Packages applied: VRTSdbcom, VRTSdbed

Build Version: 5.0.3.41 (5.0 MP3 RP4 Hotfix 1)

INSTALLING THE HOTFIX
------------------------------------
1. Stop/ Kill any dbed scripts that are running/ in use.

2. Download and copy the hotfix files to a temporary location, for example /tmp.

3. Install the hotfix patch:

	installp -d /tmp/VRTSdbcom.bff VRTSdbcom
	installp -d /tmp/VRTSdbed.bff VRTSdbed

4. Run sfua_db_config.

5. Restart vxdbd if required:

	/opt/VRTS/bin/vxdbdctrl start

UNINSTALLING THE HOTFIX
-----------------------------------
1. Stop/ Kill any dbed scripts that are running/ in use.

2. Uninstall the hotfix patch:

	installp -r VRTSdbcom 5.0.3.41
	installp -r VRTSdbed 5.0.3.41

3. Run sfua_db_config again.

4. Restart vxdbd if required:

	/opt/VRTS/bin/vxdbdctrl start

ADDITIONAL NOTES
----------------
No reboot is required after installation or un-installation of these patches.
Install these patches only on top of version 5.0 MP3 RP4.

KNOWN ISSUES (not fixed with this hotfix patch)
----------------------
Incident: 2087386
Abstract : The dbed_clonedb command for the offline checkpoint fails with message ORA-00600 with Oracle 11gR2 when ODM is enabled

Description: If ODM is enabled, the dbed_clonedb command fails when performing offline checkpoint database cloning on Oracle 11gR2.  

Sample command output:

bash-2.05b$ dbed_clonedb -S clone1 -m /clone1 -c Checkpoint_1278580468
Primary Oracle SID is FLAS11r2
New Oracle SID is clone1
Cloning an offline Storage Checkpoint Checkpoint_1278580468.
Mounting Checkpoint_1278580468 at /clone1.
All redo-log files found.
Creating initclone1.ora
        from /tmp/vxdba.rel.tmp.1515660/initFLAS11r2.ora.
Altering db_name parameter in initclone1.ora.
Editing remote_login_passwordfile in initclone1.ora.
Altering control file locations in initclone1.ora.
Copying initclone1.ora
        to /oracle/O11gr2/dbs/initclone1.ora.
About to start up new database and begin reconfiguration.
Database clone1 is being reconfigured.
Did not see error regarding database name.
Shutting down clone database.
Altering clone database archive log directory.
Updating log_archive_dest in clone database init file.
Found archive log destination at /snap_arch11r2/FLAS11r2.
Mounting clone database.
Resetting logs on new database clone1.
SFORA dbed_clonedb ERROR V-81-4920 Database clone1 is still in recovery mode.
SFORA dbed_clonedb ERROR V-81-4881 Log file is at /tmp/oralog.out.1515660.

bash-2.05b$ cat /tmp/oralog.out.1515660
ALTER DATABASE OPEN RESETLOGS
*
ERROR at line 1:
ORA-00600: internal error code, arguments: [ksfdgmsn4], [ODM ERROR V-41-4-2-207-
1 Not owner], [], [], [], [], [], [], [], [], [], []
ORA-00318: log 1 of thread 1, expected file size 512 doesn't match 512
ORA-00312: online log 1 thread 1: '/clone1/snap_data11r2/FLAS11r2/redo01.log'

Note: This issue may occur in a VVR environment.

Workaround:
Perform the offline checkpoint cloning for 11gR2 on another ORACLE_HOME where ODM is disabled.