Diagnostics Collection Script for Grid Infrastructure

Diagnostics Collection Script for Grid Infrastructure

####################################



We suggest to  use diagcollection.pl script  for collecting t diagnostic information of  Oracle Grid Infrastructure ( Oracle clusterware )

Script locate at
$GRID_HOME/bin/

/u01/app/11.2.0/grid/bin/diagcollection.pl

./diagcollection.pl --collect --all

log on as  root user then run above mention script.


When invoked script
•crsData_hostname_date.tar.gz contains log files
•coreData_host_name_date.tar.gz contains core files
•ocrData_hostname_date.tar.gz files contains ocrdump and ocrcheck also show OCR backups.
•osData_hostname_date.tar.gz contains operating system logs

To remove file from local director ..

/u01/app/11.2.0/grid/bin/diagcollection.pl –clean

OCM config & collect for single Instance

Step 1:

Remove the mentioned Old/if availabe earlier CCR directory.

   $ORACLE_HOME/ccr

Step 2:

Copy the unzip software on this location

$ORACLE_HOME/ccr

Step 3: 

Run the following from a command prompt:

    export ORACLE_HOME=/u01/app/oracle/product/11.2.0/db_1
    export ORACLE_SID=< SID NAME>

Step 4:

Go to location $ORACLE_HOME/ccr/bin then run the bellow command
   
 ./setupCCR -s CSI NUMBER email@test.com
./configCCR -s -d

Step 5:

Go to location:
 /u01/app/oracle/product/11.2.0/db_1/ccr/admin/scripts

then run the bellow command
      ./installCCRSQL.sh collectconfig -s <SID> -r sys

Step 6:
Go to location
cd $ORACLE_HOME/ccr/bin then run the bellow command

   ./emCCR collect


After collected the information, need to upload of the file ocmconfig.jar to Metalink


output location
================
/u01/app/oracle/product/11.2.0/db_1/ccr/hosts/ibdb1/state/upload/ocmconfig.jar

12.1.0.2 oracle http server showing down

12.1.0.2 oracle HTTP server showing down

 

Oracle Cloud Control 12c (12.1.0.2) OMS home HTTP service show up and down so frequently.
we found that it is bug for oracle weblogic server 11.1.
Need to apply patch "13490778 PS5:REF2: INSTANCE STATUS SHOWN INCORRECTLY IN EM WITH RC4-REFRESH-2 "and bounce OMS service.

Steps are  belows:

 Pr-Installation Instructions 
-----------------------------------

- Set the ORACLE_HOME environment variable to MW_HOME/oracle_common directory.

export ORACLE_HOME=/u01/app/oracle/product/mh/oracle_common

 Installation Instructions
--------------------------------

To install the patch, follow the instructions given in this section: 

1. Unzip the patch zip file into the PATCH_TOP. 

$unzip -d PATCH_TOP p13490778_111160_Generic.zip 

unzip -d /home/oracle/PATCH_TOP p13490778_111160_Generic.zip

2. Set your current directory to the directory where the patch is located. 

cd /home/oracle/PATCH_TOP/13490778

3. Run OPatch to apply the patch.

   /u01/app/oracle/product/mh/Oracle_WT/OPatch/opatch apply

 Post-Installation Instructions 
---------------------------------------

- Restart all servers (AdminServer and all Managed server(s)) that shares the oracle_common home.
cd /u01/app/oracle/product/mh/oms/bin 

./emctl stop oms -all
./emctl start oms 

Note: If  emctl stop oms -all  is not working, try force command 

./emctl stop oms -all -force  

 Deinstallation Instructions
-----------------------------------

If you experience any problems after installing this patch, remove the patch as follows:

1. Make sure to follow the same Pre-installation steps when deinstalling a patch. 
   This includes verifying the inventory and shutting down any services running from the 

ORACLE_HOME / machine before rolling the patch back. 

cd /u01/app/oracle/product/mh/oms/bin
./emctl stop oms -all 

 2. Change to the directory where the patch was unzipped. 

cd /export/home/oracle/PATCH_TOP/13490778 

3. Run OPatch to deinstall the patch.
    

 /u01/app/oracle/product/mh/Oracle_WT/OPatch/opatch rollback -id 13490778 

  
4. Start oms home 

./emctl start oms 

MOS Note :  1496775.1

Log Sequence in Archive Filename Wrapped

Log Sequence in Archive Filename Wrapped

Your database is running in archivelog mode and you notice the following  error sequence in the Alert file:

ARC1: Warning.  Log sequence in archive filename wrapped
to fix length as indicated by %S in LOG_ARCHIVE_FORMAT.
Old log archive with same name might be overwritten.
Sat Oct 27 13:39:51 2012

Reason behind  log_archive_format parameter value '%S'.
Explanation:

• %S Specifies the log sequence number. This number is padded to the left by zeroes. The default value is one with a range of up to five characters.

• %s Specifies the log sequence number. The number is not padded. The defaultvalue is one with no range limit on characters.  

Solution:

Modify the LOG_ARCHIVE_FORMAT parameter value in your init.ora file
to replace the uppercase %S with a lowercase %s.

Example:

Change this line in your init.ora file:
FROM:
log_archive_format = ARC_%S_%R_%T.ARC
TO:
log_archive_format = ARC_%s_%R_%T.ARC


MOS Note: 125146.1

RMAN backup at NFS mount point

RMAN backup at NFS mount point

To take rman backup at NFS mount point on Linux Platform from Oracle Database 11.2 add few rule:

Listing of required NFS options for 11.2.0.2 RMAN disk backup directory

The following NFS options must be specified for 11.2.0.2 RMAN disk backup directory:
Mount Options for Linux systems

Single Instance NFS Mount Options
---------------------------------
Mount options for Oracle Binaries

rw,bg,hard,rsize=32768,wsize=32768,vers=3,nointr,timeo=600,tcp

Mount options for Oracle Datafiles

rw,bg,hard,rsize=32768,wsize=32768,vers=3,nointr,timeo=600,tcp


RAC NFS Mount options
---------------------

Mount options for Oracle Binaries

rw,bg,hard,nointr,rsize=32768,wsize=32768,tcp, vers=3,timeo=600, actimeo=0


Mount options for Oracle Datafiles

rw,bg,hard,nointr,rsize=32768,wsize=32768,proto=tcp,noac,forcedirectio, vers=3,suid

The above values are starting recommendations as a starting point you can tune further
as required. To test which value of rsize/wsize is optimal for your system you can run
the testing outlined below.

First, we'll try transfers of 16384 blocks of 16k each from a unique file called /dev/zero
(when you read it, it's all zeroes, put out very quickly) to the test file. The time it takes to do this is our measure. From the client, command:

time dd if=/dev/zero of=/<mounted partition>/testfile bs=16k count=16384



MOS Note : ID 1117597.1