pat98's always & forever

windows에서 Opatch 진행시 oci.dll 화일 에러로 진행이 안될때가 있다.

아래와 같을 때는...

Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.
 
C:\Documents and Settings\Administrator>cd %ORACLE_HOME%\opatch
 
C:\oracle\product\10.2.0\db_1\OPatch>opatch apply C:\GTech10\p6430171_10203_WINN
T\6430171
Invoking OPatch 10.2.0.3.4
 
Oracle Interim Patch Installer version 10.2.0.3.4
Copyright (c) 2007, Oracle Corporation.  All rights reserved.
 
 
Oracle Home       : C:\oracle\product\10.2.0\db_1
Central Inventory : C:\Program Files\Oracle\Inventory
   from           : n/a
OPatch version    : 10.2.0.3.4
OUI version       : 10.2.0.3.0
OUI location      : C:\oracle\product\10.2.0\db_1\oui
Log file location : C:\oracle\product\10.2.0\db_1\cfgtoollogs\opatch\opatch2008-
04-25_17-31-16PM.log
 
ApplySession applying interim patch '6430171' to OH 'C:\oracle\product\10.2.0\db
_1'
Interim patch 6430171 is a superset of the patch(es) [  5556081 5557962 ] in OH
C:\oracle\product\10.2.0\db_1
OPatch will rollback the subset patches and apply the given patch.
 
Running prerequisite checks...
Prerequisite check "CheckActiveFilesAndExecutables" failed.
The details are:
 
Following files are active :
C:\oracle\product\10.2.0\db_1\bin\oci.dll
ApplySession failed during prerequisite checks: Prerequisite check "CheckActiveF
ilesAndExecutables" failed.
System intact, OPatch will not attempt to restore the system
 
The following warnings have occurred during OPatch execution:
1) OUI-67078:Interim patch 6430171 is a superset of the patch(es) [  5556081 555
7962 ] in OH C:\oracle\product\10.2.0\db_1
--------------------------------------------------------------------------------
 
 
OPatch failed with error code = 74
 
C:\oracle\product\10.2.0\db_1\OPatch>

-----------------------------------------------------------------------------------
원인과 해결법

Cause
This error occurs when the Oracle Files are in use.
Windows will not allow a file to be overwritten, while it is in use.

Solution
- Make sure ALL Oracle Services are stopped.
- Check to see if you have a service called Distributed Transaction Coordinator running.
- If so, stop it.
- Try to apply the patch again.

If you get the same error, then it means some other application / service is using the Oracle Files.
There is a simple workaround for this:
- Boot the machine into SAFE MODE.
- Apply the patch.
- After the patch is applied, reboot the machine back into Normal Mode.
- Finish applying the patch, by using the Post Installation Procedures.

Other possible workarounds:

Remove Oracle from the PATH, reboot the machine, apply the patch, then put Oracle back in the PATH.
Steps recommended:
- Set ALL Oracle Services to Manual Start (Note which ones are set this way, so you can
set them back later).
- Edit the System Environment Variables.
- Copy the ENTIRE PATH into a Notepad Text file (for later use).
- Remove the Oracle directories from the PATH.
- Reboot the machine.
- Apply the patch.
- Restore the PATH back to how it was (using the Notepad text file).
- Set the Oracle Services back to AUTO startup (only the ones that were set this way before).
- Reboot the machine.
- Finish applying the patch with the Post Installation Procedures (ie., the scripts).

 

It has been reported by some, an application such as a Virus Checker, Backup software, DiskKeeper, and other programs such as these, have a lock on some of the Oracle files.
You may want to disable these programs, and see if the patch gets applied.
See the next section, to help determine what process / program is locking the files.

Use a utility to determine what process is using the Oracle files.
- One such utility is called Process Explorer.
- You can find this at: http://technet.microsoft.com/en-us/sysinternals/bb896653.aspx
- Once installed, search for the Oracle .dll files.
- When you find the process that is using the files, you can kill that process.
- Then try applying the patch.

If the above steps have been tried, and you are still getting errors copying the files:
Check to see if the 'Read-Only' attribute is set, either on the patch bundle files, or>br> in the Oracle_Home.
(Note: A good indication of this problem is when there are 80+ files that can not be copied).
A common reason for this, is if the patch bundle was put onto a CD-ROM, then copied onto
the local disk. Or, if someone intentionally sets the read-only attribute on the Oracle files.

To check for the 'Read-Only' Attribute, you can:
- Use Windows Explorer
- CD to the directory (either the patch directory, or the Oracle_Home)
- Right click on a file
(Use the Opatch Apply.log, to determine the filename or folders.)
- Select Properties
- You can see if the Read-Only Attribute is set.

If the Read-Only attribute is set, you need to remove it:
- Open a CMD Prompt.
- CD to directory (the patch directory or the Oracle_Home).
- Type: attrib -r *.* /s
(This command will remove the 'read-only' attribute from all files and all sub-folders).

Now try applying the patch.

 

 

Applies to:

Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 11.2.0.2 - Release: 10.2 to 11.2
HP-UX PA-RISC (64-bit)
HP-UX Itanium
HP-UX PA-RISC - HP-UX 11iv3 September 2009 Operating Environment Update Release
HP-UX Integrity Blade Server - HP-UX 11iv3 September 2009 Operating Environment Update Release
VxFS 5.0.1 - Using OnlineJFS 5.0.1

Goal

How to use Concurrent I/O on HP-UX and improve throughput on an Oracle single-instance database.

Solution

What is Concurrent I/O ?

Concurrent I/O allows multiple processes to read from or write to the same file without blocking other read(2) or write(2) calls.

POSIX semantics requires read and write calls to be serialized on a file with other read and write calls.

With POSIX semantics, a read call either reads the data before or after the write call occurred.

With Concurrent I/O, the read and write operations are not serialized as in the case of a character device.

This advisory is generally used by applications that require high performance for accessing data and do not perform overlapping writes to the same file.

It is the responsibility of the application or the running threads to coordinate the write activities to the same file when using Concurrent I/O.

How to enable Concurrent I/O ?

Concurrent I/O can be enabled in the following ways:

A. By using the “-o cio” mount option.

The read(2) and write(2) operations occurring on all of the files in this particular filesystem will use Concurrent I/O.

- Steps for new filesystems created using 5.0.1 OnlineJFS:


- Steps for already existing filesystems which were mounted without Concurrent I/O or created with older vxfs versions:

  Existing filesystems (older filesystems created or filesystems that were not mounted with Concurrent I/O
  option) will have to be unmounted and mounted again with "-o cio" to enable Concurrent I/O.

  Note that remount command/option should not be used while mounting a filesystem with "-o cio".

1. Unmount the filesystem


2. Upgrade to VxFS 5.0.1 with 5.0.1 OnlineJFS installed on the system

Refer to “Veritas 5.0.1 Installation Guide” on http://docs.hp.com for detailed upgrade instructions.

3. Mount the filesystem with “-o cio” option


Concurrent I/O is a licensed feature of VxFS. If “-o cio” is specified, but the feature is not licensed, the mount command prints an error message and terminates the operation without mounting the filesystem.

NOTE:

Do not use "-o cio" and "-o mincache=direct,convosync=direct" together. Use either Direct I/O or Concurrent I/O.

Using Direct I/O and Concurrent I/O("-o mincache=direct,convosync=direct,cio") may cause performance regression.

B. By specifying the VX_CONCURRENT advisory flag for the file descriptor in the VX_SETCACHE ioctl command.

Only the read(2) and write(2) calls occurring through this file descriptor use concurrent I/O. The read and write operations occurring through other file descriptors for the same file will still follow the POSIX semantics.

Concurrent I/O (CIO) can be set through the file descriptor and ioctl() operation using the VX_SETCACHE ioctl command with the VX_CONCURRENT advisory flag.

The VX_CONCURRENT advisory can be set via the VX_SETCACHE ioctl descriptor on a file.

For example:


where fd is the file descriptor.

Concurrent I/O requirements

• With Concurrent I/O, the read and write operations are not serialized. It is the responsibility of the application or the running threads to coordinate the write activities and ensure they are to non-overlapping blocks of the same file.

• To gain maximum throughput, Application must perform non-overlapping writes to the same file.

• Performance increases if application write offsets are block aligned and size of I/O’s are in multiple of device block size.

• Concurrent I/O bypasses inode locking and hence application (or database used) must have its own inode-locking (serialization) mechanism for multiple writers.

• The starting file offset must be aligned to a 1024-byte boundary.

• The ending file offset must be aligned to a 1024-byte boundary, or the length must be a multiple of 1024 bytes.

NOTE:

If the Concurrent I/O alignment requirements are not met properly, then I/O’s defaults to data synchronous I/O which could cause performance degradation.

If an application issues overlapping writes(on a filesystem with Concurrent I/O enabled) to the same file without having its own serialization mechanism, behavior of write(2) is undefined and may lead to data loss.

Using Concurrent I/O with Oracle

It is recommended to place the Oracle binaries ($ORACLE_BASE directory) on a separate filesystem mounted with default options.

Placing the Oracle Database's datafiles on filesystems mounted with Concurrent I/O ("-o cio) delivers performance very close to that of raw logical volumes.

Placing Oracle binaries ($ORACLE_BASE directory) on a filesystem mounted with "cio" may cause data loss and other unexpected problems.

Hence mounting Oracle binaries ($ORACLE_BASE directory) on a filesystem mounted with "cio"option is not supported.

Concurrent I/O can significantly improve performance of a filesystem based Single instance Oracle database installation.

Concurrent I/O performs at between 93-99% of raw logical volumes.

Thus, Concurrent I/O provides performance very close to that of raw logical volumes in addition to the manageability benefits provided by a filesystem.

Concurrent IO is not expected to provide a performance benefit over direct IO when used with online and archived redo logs.

Summary

Option 1. Production and other “performance critical” databases: 3 VG’s and 5 filesystems.

	VG	Filesystem	Contents	Concurrent or Direct IO ?	Filesystem Block Size	Notes
1	vgSIDlog	origlogA	Online redo logs group A	Direct	Default	Dedicated VG can (i) improve performance as writes don’t have to wait on shared SCSI queue – as would happen if db files and logs were in same VG (ii) simplify I/O performance analysis (log writer response times).
		origlogB	Online redo logs group B	Direct	Default
2	vgSIDdata	Data1	Database files	Concurrent	8KB	Dedicated VG allows BC’s
3	vgSIDfs	Archfile	Archived redo logs	Direct	Default
		Home	Database binaries	Default	Default	Binaries could be placed on a separate VG however there is no performance benefit in doing so

Option 2. Non-production and other less performance sensitive databases; 2 VG’s and 5 filesystems.

	VG	Filesystem	Contents	Concurrent or Direct IO ?	Filesystem Block Size	Notes
1	vgSIDdata	origlogA	Online redo logs group A	Direct	Default
		origlogB	Online redo logs group B	Direct	Default
		Data1	Database files	Concurrent	8KB	If BC is used db files should be placed into a dedicated VG i.e. as per “production” layout.
2	vgSIDfs	Archfile	Archived redo logs	Direct	Default
		Home	Database binaries	Default	Default	Binaries could be placed on a separate VG however there is no performance benefit in doing so

Applies to:

Oracle Server - Enterprise Edition - Version: 11.2.0.2 and later   [Release: 11.2 and later ]
HP-UX Itanium

Symptoms

On HP-UX Itanium

After upgrade database from 11.2.0.1 to 11.2.0.2.

All V_$DIAG views are INVALID except V_$DIAG_AMS_XACTION and V_$DIAG_INFO.

ORA-942 reported on recompiling these invalid views.

ORA-942 was already reported on the same in the upgrade log file while running 'Catadrvw'.

When starting the instance, following messages are seen:
  

Error with dbgriap_init_adr_pga: 48178
   ORA-48178: error encountered while reading an ADR block file during ADR initialization [/home/oracle/oraHome112/diag/rdbms/nora/nora/metadata/ADR_INTERNAL.mif]
   ORA-48122: error with opening the ADR block file [/home/oracle/oraHome112/diag/rdbms/nora/nora/metadata/ADR_INTERNAL.mif] [0]
   ORA-27037: unable to obtain file status
   HPUX-ia64 Error: 251: Function is not available
   Additional information: 46

Changes

Upgrade to 11.2.0.2 on HP-UX Itanium

Cause

Bug 10302203: ORA-942 REPORTED WHEN COMPILING V_$DIAG VIEWS

As from 11.2.0.2, the file 'ADR_INTERNAL.mif' is opened with HP-UX directio code using
VX_SETCACHE (Set caching advisories) which is only available with the HP OnLineJFS product.

This issue happens when the HP OnLineJFS product is not installed !

Can be verified with:
    

usr/sbin/swlist -l product | grep -i vxfs

If HP OnLineJFS is installed it will give following output:
    

JFS B.11.31 Base VxFS File System 4.1 for HP-UX
OnlineJFS B.11.31 Online features of the VxFS File System

Solution

Bug 10302203 is resolved in release 12.1 and patchset 11.2.0.3 (not available at the time of creation of this article.)

A one-off patch 10302203 is available for 11.2.0.2 on HP-UX Itanium and can be downloaded at MOS.

Workarounds:

1. Install the HP OnLineJFS product

-OR-

2. Set init.ora parameter _DIAG_ADR_ENABLED to FALSE

    ADR is disabled and all tracing goes to $ORACLE_HOME/rdbms/log

In this Document
  Purpose
  Pre 11.2 Database Issues in 11gR2 Grid Infrastructure Environment
  References

---

Applies to:

Oracle Server - Enterprise Edition - Version: 10.1.0.2 to 11.1.0.8 - Release: 10.1 to 11.1
Information in this document applies to any platform.

Purpose

The note lists most known issues in 11gR2 Grid Infrastructure (in short GI) + pre-11gR2 database environment.

Even workaround is available in some cases, it's recommended to apply patches whenever possible. Refer to note 1064804.1 for instructions to patch in mixed environment.

For CRS PSU/bundle patch information, refer to note 405820.1 for 10.2 and note 810663.1 for 11.1

Pre 11.2 Database Issues in 11gR2 Grid Infrastructure Environment

1. Error creating or starting pre-11.2 database:

If it happens while creating database, DBCA fails with ORA-29702 and traces shows:



If it happens while starting existing database, sqlplus startup fails:



Database alert.log shows the following message:

Solution:

To start a pre11gR2 database in 11gR2 Grid Infrastructure environment, node(s) must be pinned. To pin node(s), as root execute:

To find out whether node(s) is pinned or not:

2. If datafiles are located in ASM, DBCA fails to create database with error: 

DBCA trace (10g in $RDBMS_HOME/cfgtoollogs/dbca and 11g in $ORACLE_BASE/cfgtools/dbca) shows the following exception:


Solution:

Due to unpublished bug 8288940 (fixed in 10.2.0.5), DBCA will fail if database files are located in ASM. patch 8288940 is platform independent and is available for 10.2.0.3, 10.2.0.4, 11.1.0.6 and 11.1.0.7 as a .jar file and needs to be applied to database home. For other RDBMS version where there's no patch, please refer to workaround section of bug 8520511 in Database Readme <Oracle Database Readme 11g Release 2 (11.2) -> Bug 8520511>

3. SRVCTL fails to start instance if OCR is located in an ASM diskgroup or with different permission/ownership. 

The racgimon log (located in $RDBMS_HOME/log/$HOST/racg/imon_${DBNAME}.log) shows the following message:

Solution:

Due to unpublished bug 8262786, if OCR is located in ASM or with different permission/ownership, srvctl will fail to start earlier database version.Fix for unpublished bug 8262786 is included in 10.2.0.4 CRS PSU4, 10.2.0.5, 11.1.0.7 CRS PSU4, and Windows 10.2.0.4 Patch 36 and needs to be applied to database home. The workaround is to use sqlplus to start the database instead of srvctl

@ 8312004 closed as dup

4. Database fails to start after restart of GI.

$GRID_HOME/log/$HOST/agent/crsd/application_<dbuser>/application_<dbuser>.log shows:

After GI restart, status of diskgroup:


Solution:

Due to unpublished bug 8448079, while stopping GI, ASM init.ora parameter asm_diskgroups will be nullified and some diskgroups will remain OFFLINE after restart of GI which cause pre-11.2 database fails to start. Fix for unpublished bug 8448079 is included in the 11.2.0.2, patch 8448079 exists for certain platform and needs to be applied to GI home.  The workaround is to add dependence of diskgroup to each instance:




5. SRVCTL fails to start service

For example:


Solution:

Due to unpublished bug 8373758, pre-11.2 srvctl will fail to start a service if the service is followed by a 11gR2 new resource in "crsctl stat res" output.  In the above example, ora.eons is a new resource in 11.2 and pre-11.2 srvctl can't parse its status properly. Fix for unpublished bug 8373758 is included in 10.2.0.4 CRS PSU4, 10.2.0.5, 11.1.0.7 CRS PSU2 and needs to be applied to database home.  The workaround is to:

And execution of setasmgidwrap fails with:



Solution:

Due to bug 9575578, setasmgidwrap fails with pre-11.2 oracle binary. Fix for bug 9575578 is included in 11.2.0.2, patch 9575578 exists for certain platform and needs to be applied to GI home.

7. After removal of pre-11.2 CRS home, the following error reported while trying to start or stop database, or stop cluster:

If GI is being stopped, the following will be reported:



Solution:

Due to bug 9257105, even upgrade finishes successfully, OCR configuration for pre-11.2 database still points to pre-11.2 CRS home. Fix for bug 9257105 is included in 11.2.0.1.2 and 11.2.0.2, unfortunately the fix itself is having regression which is being worked in unpublished bug 9678856. The workaround is to:



8. Singleton service does not failover or uniform service does not stop after local node VIP resource failed or stopped:


Solution:

Due to unpublished bug 9039498, pre-11.2 database service will not failover or stop if local public network is down. Fix for bug 9039498 is included in 11.2.0.2, 12.1 and applies to GI home.



9. DBCA/srvctl fails to add instance/database with the following error:


Solution:

Due to bug 9767810, if pre-11.2 database is not installed on all nodes of the cluster, srvctl fails to add instance/database to OCR. bug 9767810 is fixed in 11.2.0.1.2 and applies to GI home. The workaround is to copy pre-11.2db/bin/racgwrap to all nodes in the cluster, and make sure it's accessible by pre-11.2 database owner.

10. Pre-11.2.0.2 database fails if any private network fails.

It happens as pre-11.2.0.2 instance isn't aware of Redundant Interconnect feature (multiple active cluster_interconnect in "oifcfg getif"), see note 1210883.1 for more about HAIP

Solution:

As HAIP feature can not be disabled, for environment with GI 11.2.0.2 with pre-11.2.0.2 database/ASM, it's recommended to use OS level bonding solution for private network as in earlier clusterware version.

11. If OCR is located on ASM, pre-10.2.0.5 database can not get cluster interconnect even it's configured in OCR:

Solution:

For 11gR2 GI + pre-10.2.0.5 database environment, it's not recommended to place OCR on ASM. Other workaround is to set init.ora parameter cluster_interconnects and ignore the warning.

@ 9865139 closed as duplicate of 5389506 which is only fixed in 10.2.0.5 and above

12. After clusterware upgrade, ocrdump shows key "SYSTEM.ORA_CRS_HOME" is not updated to new clusterware home.

If previous clusterware home is renamed or removed, pre-11gR2 SRVM client (srvctl, DBCA, DBUA etc) fails.

In this example, /ocw/b201 is pre-upgrade clusterware home which is not updated.

Solution:

Due to bug 10231584, OCR key SYSTEM.ORA_CRS_HOME is not updated during upgrade. The workaround is to execute the following as root on any node:




13. Service does not stop/failover after stopping corresponding instance

$DB_HOME/log/$HOST/racg/imon_<DB_NAME>.log

2011-03-14 23:22:52.228: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: CLSR-0521: Event ora.ha.racdb.racdb1.inst.down is rejected by EVM daemon
2011-03-14 23:22:52.228: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: clsrcepevm: clsrcepostevt status = 17

2011-03-14 23:22:52.228: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: clsrcep:evm post return 1
2011-03-14 23:22:54.458: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: CLSR-0521: Event sys.ora.clu.crs.app.trigger is rejected by EVM daemon

2011-03-14 23:23:06.495: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: clsrcexecut: env _USR_ORA_PFILE=/ocw/grid/racg/tmp/ora.racdb.racdb1.inst.ora

2011-03-14 23:23:06.495: [    RACG][1108842816] [14693][1108842816][ora.SD302.SD3021.inst]: clsrcexecut: cmd = /database/db205/bin/racgeut -e _USR_ORA_DEBUG=0 -e ORACLE_SID=racdb1 540 /database/db205/bin/racgimon stop ora.racdb.racdb1.inst

$GRID_HOME/log/$HOST/evmd/evmd.log

2011-02-16 06:13:55.965: [  EVMAPP][4163668704] EVMD Started
..
2011-02-16 06:13:55.980: [    EVMD][4163668704] Could not open /ocw/grid/evm/admin/conf/evmdaemon.conf Reconfiguration aborted.

Solution:

Due to bug 12340700, the following files has wrong permission:

ls -l $GRID_HOME/evm/admin/conf
-rw------- 1 root root 3032 Feb 19 14:42 evm.auth
-rw------- 1 root root 2318 Feb 19 14:42 evmdaemon.conf
-rw------- 1 root root 4871 Feb 19 14:42 evmlogger.conf

Unpublished bug 12340700 is fixed in 11.2.0.3 and applies to GI home. The workaround is to fix the permission manually with "chmod" command and restart GI. The expected ownership and permission are:

ls -l $GRID_HOME/evm/admin/conf
-rw-r--r-- 1 root root 3032 Feb 19 14:42 evm.auth
-rw-r--r-- 1 root root 2318 Feb 19 14:42 evmdaemon.conf
-rw-r--r-- 1 root root 4871 Feb 19 14:42 evmlogger.conf

14. 10.1 ASM or DB does not start during or after clusterware upgrade

During upgrade, OUI shows the following error:

[INS-20802] Automatic Storage Management Configuration Assistant failed.

Installation log ($ORACLE_BASE/oraInventory/logs/installActions{$TIMESTAMP}.log):

INFO: Read: PRCR-1079 : Failed to start resource ora.asm
INFO: Read: CRS-2672: Attempting to start 'ora.asm' on 'racnode2'
INFO: Read: CRS-5017: The resource action "ora.asm start" encountered the following error:
INFO: Read: ORA-29702: error occurred in Cluster Group Service operation
INFO: Read: . For details refer to "(:CLSN00107:)" in "/ocw/grid/log/racnode2/agent/ohasd/oraagent_oracle/oraagent_oracle.log".
INFO: Read:
INFO: Read: CRS-2674: Start of 'ora.asm' on 'racnode2' failed

Solution:

This is caused by bug 3841387, the solution is to apply patch 9632335 to 10.1 ASM and DB home and retry. Refer to note 969254.1 Section "OUI Plug-in Failure" for retry steps, alternately ASMCA can't be executed manually to upgrade ASM.


15. pre-11.2 instance/database stops after check timed out:

$GRID_HOME/log/<node>/crsd/crsd.log



$GRID_HOME/log/<node>/agent/crsd/application_<dbuser>/application_<dbuser>.log



$DB_HOME/log/<node>/racg/imon_<dbname>.log



Instance alert.log

Solution:

Due to bug 11933693, pre-11.2 database/instance will be stopped if agent check times out. The bug will not cause timeout, the cause could be slow/hang OS/DB and needs to be investigated separately.

The bug is fixed in 11.2.0.3 and applies to GI home.

If you get the following errors whilst starting up nodeapps on a 10G RAC cluster

bart-> srvctl start nodeapps -n bart
CRS-1028: Dependency analysis failed because of:
CRS-0223: Resource 'ora.bart.gsd' has placement error.
CRS-1028: Dependency analysis failed because of:
CRS-0223: Resource 'ora.bart.ons' has placement error.

Check the status of these resources.

bart-> ./ShowCRS | grep UNK
ora.bart.gsd ONLINE UNKNOWN on bart
ora.bart.ons ONLINE UNKNOWN on bart

To clear the state of the resource you have to use the force option
in the crs_stop command

bart-> crs_stop ora.bart.gsd -f
Attempting to stop `ora.bart.gsd` on member `bart`
Stop of `ora.bart.gsd` on member `bart` succeeded.

bart-> crs_stop ora.bart.ons -f
Attempting to stop `ora.bart.ons` on member `bart`
Stop of `ora.bart.ons` on member `bart` succeeded.

bart-> srvctl start nodeapps -n bart
bart-> ./ShowCRS | grep UNK

bart-> ./ShowCRS | grep bart
ora.bart.gsd ONLINE ONLINE on bart
ora.bart.ons ONLINE ONLINE on bart
ora.bart.vip ONLINE ONLINE on bart

ShowCRS is a shell script which is based on Metalink Note 259301.1

--------------------------- Begin Shell Script -------------------------------

#!/usr/bin/ksh
#
# Sample 10g CRS resource status query script
#
# Description:
#    - Returns formatted version of crs_stat -t, in tabular
#      format, with the complete rsc names and filtering keywords
#   - The argument, $RSC_KEY, is optional and if passed to the script, will
#     limit the output to HA resources whose names match $RSC_KEY.
# Requirements:
#   - $ORA_CRS_HOME should be set in your environment

RSC_KEY=$1
QSTAT=-u
AWK=/usr/xpg4/bin/awk    # if not available use /usr/bin/awk

# Table header:echo ""
$AWK \
  'BEGIN {printf "%-45s %-10s %-18s\n", "HA Resource", "Target", "State";
          printf "%-45s %-10s %-18s\n", "-----------", "------", "-----";}'

# Table body:
$ORA_CRS_HOME/bin/crs_stat $QSTAT | $AWK \
'BEGIN { FS="="; state = 0; }
  $1~/NAME/ && $2~/'$RSC_KEY'/ {appname = $2; state=1};
  state == 0 {next;}
  $1~/TARGET/ && state == 1 {apptarget = $2; state=2;}
  $1~/STATE/ && state == 2 {appstate = $2; state=3;}
  state == 3 {printf "%-45s %-10s %-18s\n", appname, apptarget, appstate; state=0;}'

--------------------------- End Shell Script -------------------------------

Metalink (문서 ID 749160.1)

node1을 stop 했는데 VIP가 넘어가지 않을때, 10.2.0.5 로 patch or 11g 사용권장
========================================================================================

Running a 2 node rac with 10.2.0.3. When nodeapps are stopped from one node, the vip-interface
does not failover to existing nodes. Therefore all users can not connect to the database.
When issuing command :

then VIP-interface of node node1 does not get migrated to node node2.

As you entered the stop command manually this is interpreted by crs not as a failure but as a wish
from your side and so no failover takes place.

The behavior has changed in 11g and will in 10.2.0.5 due to enhancement request 4623630.

 In 11g and 10.2.0.5 you will use the following syntax:

The -r switch tells the system to relocate the VIP to an available node.

In case this functionality is important, the best is to upgrade to 11g or 10.2.0.5 as it becomes available.

1. REDO
 - 모든 변경사항을 REDO로그에 기록한다.
 - REDO로그는 Online REDO와 Archuved REDO로그로 구성된다.
    Online REDO 로그는 2개 이상의 파일로 구성되어 있어,
    현재 사용중인 로그 파일이 꽉 차면 다음 로그 파일로 스위칭
    이때 꽉 차여진 로그 파일을 다른 위치로 백업래 준 파일을 Archived REDO로그 이다.
 - 목적 3가지
  - 데이타 복구 : Archived REDO이용
  - 버퍼캐시복구 : 인스턴스가 비정상적종료시 그떄까지 작업내용이 잃어버리게 됨.
                          재기동 되면 Online REDO로그에 저장된 기록사항을 읽어와 마지막
                          채크포인트와 사고발생직전까지 수행한 트랜젹션을 재현
  - Fast Commit(IO속도 차 극복) : 데이타 버퍼 블록을 디스크에 기록하는 작업은 Random엑세스 방식,
                                              Append방식 Append방식이 상대적으로 빠르게 때문에 우선 변경사항을
                                              Append방식으로 기록하고 동기화는 후에 배치방식으로 일괄수행.
 - REDO로그 버퍼를 REDO로그에 기록하는 시점
  - 3초마다 DBWR프로세스부터 신호흫 받을때
  - 로그 버퍼의 1/3이 차거나 기록된 REDO레코드량이 1MB를 넘을떄
  - 사용자 커킷 또는 롤백 명령이 날릴때
  
2. UNDO
 - 각 트랜지션별로 UNDO세그먼트를 할당해주고 그 트랜지션이 발생시킨 테이블과 인댁스에
    대한 변경사항을 UNDO레코드 단위로 UNDO세그먼트 블록에 기록.
 - 목적 3가지
  - 트랜지션 롤백
  - 트랜지션 리커버리(인스턴스 리커버리시 롤백단계)
  - READ Consistency(읽기 일관성)
  타 DBMS는 Lock를 통해 일기 일관성을 구현하지만,
  오라클에서는 UNDO데이타를 이용해서 읽기 일관성을 구현한다.  
             읽기 일관성이란 Transaction이 진행되는 동안 Database의 다른 사용자는 이 Consistent Read에 의해
             Commit되지 않은 변경 사항을 볼 수 없는 기능 입니다.  
 - UNDO레코드에 기록되는 내용
  - Insert : 추가된 레코드의 rowid
  - Update : 변경되는 컬럼에 대한 before image
  - Delete : 지워지는 로우의 모든 컬럼의 대한 before image
------------------------------------------------------------------------------------
REDO 와 UNDO를 차이점에서 바라본 관점.
------------------------------------------------------------------------------------
REDO 는 UNDO를 포함 합니다.
REDO 는 시스템 장애시 복구를 위해 사용 합니다.
복구시에 UNDO 데이터도 같이 복구하구요. Commit 되지 않은 데이터를 Rollback 하게 됩니다.
UNDO 는 Rollback 시에도 사용 되지만 Read Consistency(읽기 일관성) 을 위해서 도 사용 됩니다.
REDO 는 모든 변경사항(UNDO 포함)을 기록 합니다.
복구는 UNDO 를 통해서 복구를 하게 됩니다. 즉, ROLLBACK을 한다는 말이죠.
시스템 장애가 발생하게 되면 UNDO 데이터도 모두 날아가게 되겠죠.
결국 시스템 장애시 REDO 데이터를 이용해서 마지막 CHECK POINT 부터 장애까지의 DB BUFFER CACHE 를 복구하게 됩니다.
이게 완료가 되면 UNDO DATA 를 이용하여 COMMIT 되지 않은 데이터를 모두 ROLLBACK 함으로써 복구를 완료하게 됩니다.
결국 REDO 가 UNDO 를 복구하고 최종적으로 UNDO가 복구를 하게 됩니다.
UNDO(안한것 처럼)는 되돌리는 것 이라고 보시면 될거에요.
어떤 세션에서 DML을 발생시키면 commit이나 rollback을 날리기 전까지 이전 정보를 저장하기 위해서
UNDO 블럭에 해당 정보를 기록하죠.. 해당 세션이 트렌젝션 중에 비정상적으로 종료가 되었다면,
안한것 처럼(UNDO) 다시 원복해야 되니까요.
REDO(다시 함)는 위에 설명에도 있듯이 인스턴스 실패시(DB가 내려감) 데이터 파일에 쓰여지지 않은 커밋된 데이터를 복구한다고 되어 있습니다.
좀 더 설명을 해드리자면 커밋을 날릴 경우 LGWR가 로그 파일에 변경된 정보를 기록하게 되고
그 다음에 DBWR이 버퍼캐쉬에 있는 변경 사항에 대해서 데이터 파일에 내려 쓰게 됩니다.
이는 복구 정보가 더 중요해서 그런 것이라고 이해하시면 됩니다(복구 정보가 있으면 재적용이나 되돌릴 수 있으므로).
DB가 비정상적으로 내려가면, 데이터 파일에 쓰여지지 않은 정보들이 리두로그에 기록되어 있으므로
리두로그에서 정보를 가져와 다시 적용(REDO)하는 것입니다.

Applies to:

Oracle Server - Enterprise Edition - Version: 10.2.0.5 and later   [Release: 10.2 and later ]
Microsoft Windows (32-bit)
Microsoft Windows x64 (64-bit)
Microsoft Windows x64 (64-bit) - OS Version: 7
Microsoft Windows (32-bit) - OS Version: 7
Microsoft Windows x64 (64-bit) - Version: 2008 R2

Goal

To explain how to install 10.2.0.5 Oracle Client software on Windows 7 and Windows 2008 R2.

NOTE:  Oracle 10g R2 (10.2) Client Software was NOT supported on Windows 7 (32/64 bit) or Windows 2008 R2 (32/64 bit) when it was initially released.  These OS's were not available at that time.  Oracle decided to start supporting them with the release of the 10.2.0.5.0 patch.  Please review the following Statement of Direction for more details:

Note 1061272.1 Statement of Direction: Oracle Database 10g Release 2 Client (10.2.0.5) with Microsoft Windows 7 and Windows Server 2008 R2

Solution

How To Install 32-bit 10.2.0.5 Oracle Client Software on Windows 7 (32 or 64-bit) or Windows 2008 R2 (32 or 64-bit)

1. Download the following base software from TechNet:
    
        Oracle Database 10g Client Release 2 (10.2.0.3)
        10203_vista_w2k8_x86_production_client.zip (616,757,461 bytes)
   
   
2. Extract the software.  Browse the extracted files and find setup.exe which is located inside the Disk1 folder.
   
   
3. Open a Command Prompt, change to the location where setup.exe resides and issue the following command:
   
        C:\Temp\10203\client\Disk1> setup.exe -ignoreSysprereqs -ignorePrereq
   
   
   NOTE:  The parameters are CASE-SENSITIVE and must be entered EXACTLY as shown above.
   
   
   These flags attempt to suppress any Windows 7/2008 R2 warnings from being raised about the OS not being compatible with the Oracle Universal Installer (OUI).  Even if warnings are raised when running the OUI you will still be able to proceed with the software installation (i.e. click the Next button).
   
   
4. After the software installation completes, you must download the 10.2.0.5 patch and apply it.
   
   
   1. Log into My Oracle Support.
      
      
   2. Click on the Patches & Updates tab.
      
      
   3. In the Patch Search section set Patch Name or Number to 8202632.
      
      
   4. Set the Platform to Microsoft Windows (32-bit).
      
      
   5. Click Search.
      
      
   6. Download the following patch:
      
            Patch 8202632: 10.2.0.5.0 PATCH SET FOR ORACLE DATABASE SERVER
      
      
      NOTE:  Do not be confused by the name of the patch containing the words Oracle Database Server.  This patch is used to upgrade BOTH Oracle Client AND Oracle Database software.
      
      
      
   7. Once the patch has been downloaded, extract the patch.  Browse the extracted files and run setup.exe, which is located inside the Disk1 folder, to install the 10.2.0.5 patch.
      
      
      NOTE:  When the OUI asks you to Specify Home Details be sure the choose your EXISTING 10.2.0.3 Oracle Home installed in Step 3.  If you do not see the correct Name and Path to your current Oracle Home, click the drop-down button and choose it from the list.  DO NOT attempt to install this patch into a NEW Oracle Home.
      
      
      
   
   
5. Once the patch has been applied you will be running a supported version of 32-bit 10g R2 client software on a Windows 7/Windows 2008 R2 OS.
   

How To Install 64-bit 10.2.0.5 Oracle Client Software on Windows 7 (64-bit) or Windows 2008 R2 (64-bit)

1. Download the following base software from TechNet:
    
        Oracle Database 10g Client Release 2 (10.2.0.4)
        10204_vista_w2k8_x64_production_client.zip (507,934,134 bytes)
   
   
2. Extract the software.  Browse the extracted files and find setup.exe which is located inside the Disk1 folder.
   
   
3. Open a Command Prompt, change to the location where setup.exe resides and issue the following command:
   
        C:\Temp\10204\client\Disk1> setup.exe -ignoreSysprereqs -ignorePrereq
   
   
   NOTE:  The parameters are CASE-SENSITIVE and must be entered EXACTLY as shown above.
   
   
   These flags attempt to suppress any Windows 7/2008 R2 warnings from being raised about the OS not being compatible with the Oracle Universal Installer (OUI).  Even if warnings are raised when running the OUI you will still be able to proceed with the software installation (i.e. click the Next button).
   
   
4. After the software installation completes, you must download the 10.2.0.5 patch and apply it.
   
   
   1. Log into My Oracle Support.
      
      
   2. Click on the Patches & Updates tab.
      
      
   3. In the Patch Search section set Patch Name or Number to 8202632.
      
      
   4. Set the Platform to Microsoft Windows x64 (64-bit).
      
      
   5. Click Search.
      
      
   6. Download the following patch:
      
            Patch 8202632: 10.2.0.5.0 PATCH SET FOR ORACLE DATABASE SERVER
      
      
      NOTE:  Do not be confused by the name of the patch containing the words Oracle Database Server.  This patch is used to upgrade BOTH Oracle Client AND Oracle Database software.
      
      
      
   7. Once the patch has been downloaded, extract the patch.  Browse the extracted files and run setup.exe, which is located inside the Disk1 folder, to install the 10.2.0.5 patch.
      
      
      NOTE:  When the OUI asks you to Specify Home Details be sure the choose your EXISTING 10.2.0.4 Oracle Home installed in Step 3.  If you do not see the correct Name and Path to your current Oracle Home, click the drop-down button and choose it from the list.  DO NOT attempt to install this patch into a NEW Oracle Home.
      
      
      
   
   
5. Once the patch has been applied you will be running a supported version of 64-bit 10g R2 client software on a Windows 7/Windows 2008 R2 OS.
   

---

Applies to:

Oracle Server - Enterprise Edition - Version: 11.1.0.6 to 11.2.0.2 - Release: 11.1 to 11.2
Information in this document applies to any platform.
Oracle Server Enterprise Edition - Version: 11.1 to 11.1
***Checked for relevance on 14-Jan-2011***

Goal

The purpose of this document is to provide information about the Copy the asm files between remote instances and local instances using ASMCMD.

Solution

11g new feature,you can use asmcmd to copy files between remote instances
Enables you to copy files between ASM disk groups on local instances and remote instances. You can also use this command to copy files from ASM disk groups to the operating system.

cp -ifr <source file name> <user_name>@<host_name>.<Port Number>.<SID>:<targer path>/<target file name>

user_name@host_name.<Port Number>.<SID>

The user_name, host_name, and SID are required. The default port number is 1521.

Troubleshooting-ASMCMD remote copy

asmcmd remote copy works through listener connection.

ASMCMD remote connection can fail with below generic error.

ASMCMD-08202: internal error: [asmcmdshare_error_msg_05] [8201]

[8201] means unable to connect remote ASM Instance.

It could be due to following reason.

We need to enable additional tracing for asmcmd connection to get a exact failure message.

++ set DBI_TRACE environment variable for asmcmd perl tracing.

export DBI_TRACE=1

++ Now connect using asmcmd and re-produce the issue.

Example 1:-

Here we can see that asmcmd copy failed due to ORA-01031.

ASMCMD uses SYSASM by default if -a option is not used.

Here the problem is sysasm privelege was not given to sys user on remote ASM Instance.

Given the SYSASM privilege to SYS ( or the user trying to connect ). When you grant a system privilege, the password file is updated.


SQL> grant sysasm to sys;
Grant succeeded.

SQL> select * from v$pwfile_users;
USERNAME SYSDB SYSOP SYSAS
------------------------------ ----- ----- -----
SYS TRUE TRUE TRUE

Now the remote asmcmd copy works fine

For more detals,please go through the below notes

Note.730067.1 - Troubleshooting ORA-1031 Insufficient Privilege
Note.578796.1 - ORA-01031 While Connecting as SYSASM

Example 2:-

Solution:-

The cp command failed because the target ASM file name was not specified  or File name should not contain the file number/incarnation.We can not copy OMF form files without specifying file name

In this Document
  Purpose
  Scope and Application
  11gR2 RAC TAF Configuration for Admin and Policy Managed Databases
     Concepts & Overview
     Administrator-managed Database
     Policy-managed Database
  References

---

Applies to:

Oracle Net Services - Version: 11.2.0.1 to 11.2.0.2 - Release: 11.2 to 11.2
Oracle Server - Enterprise Edition - Version: 11.2.0.1 to 11.2.0.2   [Release: 11.2 to 11.2]
Information in this document applies to any platform.

Purpose

This article discussed about Transparent Application Failover (TAF) and its configuration details for 11gR2 cluster database.

Scope and Application

This article is intended for Oracle Cluster Administrators who would like more details about TAF. Basic understanding of Oracle Clusterware Database is assumed.

11gR2 RAC TAF Configuration for Admin and Policy Managed Databases

Concepts & Overview

TAF allows Oracle clients to reconnect to surviving instance in the event of a failure of the current instance to which it connected. There are two types of TAF available, SESSION and SELECT.

SESSION: Session Failover re-creates the connections and sessions to the surviving instance.
SELECT:  In addition to recreating the session, Select Failover also replays the queries that were in progress.


There are two method in which TAF estabilishes the faiover connection, BASIC and PRECONNECT.

BASIC: The second connection is re-established only after the first connection to the instance failed.
PRECONNECT: Two connections are established when first login to database. A login to database will create two connection at the same time. For this to work clusterware actually starts two service. One main service and another shadow service.


TAF can be configured at the Oracle Client side in tnsnames.ora or at the Database Server side using srvctl. Configuring it at the server is preferred as it is convenient to put the configuration in one place i.e. on the server. This article focuses on configuration at the server using srvctl. TAF configuration is associated with the database service. Creating the service with TAF related options is all about configuring TAF.

1. Create service with srvctl
srvctl defines services in OCR which is accessible across nodes. The service configuration details can be viewed with srvctl config service command.

2. Start the Service with srvctl
When the service is started with srvctl, it actually creates a service in the database i.e. RDBMS. The service name and the configuration details of the service created at the database can be viwed with DBA_SERVICES view. Then the SERVICE_NAMES database parameter is set, which allows PMON to register this service with the listener. Hence connections can be established through listener to the database.



Let's have a look at the srvctl syntax for creating service. Databases in 11gR2 can be Administrator-managed or Policy-managed. Certain options can be used only with certain type of database.


Here is the description of the options that we are going to use in configuring TAF.

-r "<preferred_list>" [-a "<available_list>"] [-P {BASIC | NONE | PRECONNECT}]
         This clause is valid only for Administrator-managed database which support PRECONNECT method.
-g <server_pool> [-c {UNIFORM | SINGLETON}]
         This clause is valid only for Policy-managed database where in PRECONNECT method is not available.

The above two options instruct clusterware on how to handle this service.

[-e {NONE|SESSION|SELECT}]
        This defines the type of TAF whether SESSION or SELECT.
[-m {NONE|BASIC}]
        This defines the method of TAF.
[-z <failover_retries>]
        This defines the the number of times to attempt to connect after a failover.
[-w <failover_delay>]
       This defines the amount of time in seconds to wait between connect attempts.

Above four options are passed to database while starting up the service and is viewable in DBA_SERVICES view.

Administrator-managed Database

Basic Method

1. Create Service

Syntax:
          srvctl add service -d <db_unique_name> -s <service_name>
          -r "<preferred_list>" [-a "<available_list>"] [-P {BASIC | NONE | PRECONNECT}]
         [-e {NONE|SESSION|SELECT}]
         [-m {NONE|BASIC}]
         [-z <failover_retries>]
         [-w <failover_delay>]


Example:
         $ srvctl add service -d db112 -s mysrvb -r db1121,db1122 -P basic -e select -m basic -z 10 -w 2



2. Start the service
$ srvctl start service -d db112 -s mysrvb

Check if the service is rightly registered with the listener.



3. Make a connection.

The client connect string do not require any special TAF parameters. A simple tnsnames.ora entry as shown below will do.


$ sqlplus scott/tiger@mysrvb

Once connectd GV$SESSION view will reflect the connection made is TAF enabled. The FAILOVER_METHOD and FAILOVER_TYPE column reflects it and this confirms the TAF configuration is correct.


Note that this has connected to the instance 1 with BASIC method and SELECT type.

Preconnect Method

1. Create service

Syntax:
           srvctl add service -d <db_unique_name> -s <service_name>
          -r "<preferred_list>" [-a "<available_list>"] [-P {BASIC | NONE | PRECONNECT}]
          [-e {NONE|SESSION|SELECT}]
          [-m {NONE|BASIC}]


Example:
          $ srvctl add service -d db112 -s mysrv -r db1121 -a db1122 -P preconnect



2. Start the service.

$ srvctl start service -d db112 -s mysrv

Preconnect creates an additional shadow service.



Note that the service 'mysrv' is started on the preferred instance. And the shadow service 'mysrv_preconnect' is started on the available instance. The shadow service is always defined as <servicename>_preconnect.


Check if the service is correctly registered with the listener.




3. Define TNS entry.

Preconnect requires special parameter at the client side tnsnames.ora file. The TAF Type and the TAF method is actually specified at the client end.



Connect to the database with this tns entry

$sqlplus scott/tiger@mysrv


Note that it has made two connections. The first one is the active main connection with PRECONNECT method and SELECT type. The second one is the shadow connection that will be used when the connection to the first instance fails.

Policy-managed Database

Basic Method

1. Create the service

Syntax:
          srvctl add service -d <db_unique_name> -s <service_name>
          -g <server_pool> [-c {UNIFORM | SINGLETON}]
          [-e {NONE|SESSION|SELECT}]
          [-m {NONE|BASIC}]
          [-z <failover_retries>]
          [-w <failover_delay>]


Example:
         $ srvctl add service -d db1 -s mysrvp -g mysrvpool -c uniform -e select -m basic -z 10 -w 2


2. Start the service

$ srvctl start service -d db1 -s mysrvp


Check if the service is correctly registered with the listener



3. Make a connection.

The client connect string do not require any special TAF parameters. A simple tnsnames.ora entry as shown below will do.


$ sqlplus scott/tiger@mysrvp