You want to participate in Oracle Beta program or Exam?

Oracle Technology Beta programs available for all Oracle technology products including Oracle Database, Oracle Enterprise manager, Oracle Fusion Middleware…etc.

Please note that Beta program is highly selective and acceptance into these programs makes you part of an elite group of test sites that are carefully selected based up on technical expertise and knowledge of Oracle productions

Why should you participate?

The main benefit is gaining access to technology before your competitors and developing a close working relationship with oracle development.

If you want to participate in Oracle beta program, please register yourself using below link.
http://otnbeta.oracle.com/bpo/prospects/index.htm

You can also contact BETAPROG_US@ORACLE.COM for details

You can also participate in Oracle Certification Beta exams. These Beta exams are pre-production exams used by Oracle to evaluate new exam questions and it will go through a rigorous development and evaluation process before being released to the public.

Beta exams are offered at a deeply discounted price and exam usually takes around 3 to 3.5 hours to complete.

To learn more information about Beta Exams, please visit
http://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=182

Regards,
Satishbabu Gunukula
http://www.oracleracepxert.com

How to STOP and START processes in Oracle RAC and Log Directory Structure

You need to follow the below steps to Stop and Start processes in Oracle RAC

Shutdown RAC Database
You need to Shutdown Database instances on each node. You can either use Oracle Enterprise Manager or SVRCTL to shutdown the instances. If you are using EM Grid control then set a blackout in Grid control for processes that you intend to shutdown. So that records for these processes indicate that the shutdown was planned.
Use below command to stop Enterprise Manager/Grid Control
$ORACLE_HOME/bin/emctl stop dbconsole

Use below command to shutdown all oracle RAC instances on all nodes.
$ ORACLE_HOME/bin/srvctl stop database -d db_name

If you want to stop specific database instances use below command
$ ORACLE_HOME/bin/srvctl stop database -d db_name –i instance_name

Shutdown Oracle ASM Instance
Once the database is stopped, proceed with ASM Instance shutdown.
Use below command to shutdown ASM instances on all nodes
$ORACLE_HOME/bin/bin/srvctl stop asm -n node

Shutdown Node applications
Use below command to shutdown node apps on all RAC nodes
$ORACLE_HOME/bin/bin/srvctl stop nodeapps -n node

Shutdown Oracle Clusterware
You need to Shutdown oracle clusterware or CRS as root and run below command on each node in the cluster.
#crsctl stop crs

Please note that using above command will stop Oracle High availability services (OHAS) and Clustware stack in a single command

From 11g R2, you can do this in two stops

1. Stop Clustwerware stack on local node
#crsctl stop cluster

You can stop the clusterware stack on all nodes in the cluster
# Crsctl stop cluster –all

Where
-all Start clusterware on all nodes
-n Start clusterware on particular nodes

2. Stop Oracle High availability service demon on each node in the cluster.
# crsctl stop has

Check the Status of Cluster
Once all process stopped run the below command to check the status of CRSD,CSSD,EVMD process.
# crsctl check crs

If you see any process failed to stop then you can also use Force option to terminate the processes unconditionally.
$ crsctl stop crs –all –f

Start processes in Oracle RAC
Follow the reverse sequence to start all processes in oracle RAC
# crsctl start crs
$ORACLE_HOME/bin/bin/srvctl start nodeapps -n node
$ORACLE_HOME/bin/bin/srvctl start asm -n node
$ORACLE_HOME/bin/srvctl start database -d db_name

If you come across any issues during startup orshutdown, check the Oracle Clusterware Component Log Files.

Oracle Clusterware Log Directory Structure

CRS_HOME/log/hostname/crsd/ - The log files for the CRS daemon CRS_HOME/log/hostname/cssd/ - The log files for the CSS daemon
CRS_HOME/log/hostname/evmd/ - The log files for the EVM daemon
CRS_HOME/log/hostname/client/ - The log files for the Oracle Cluster Registry (OCR)
CRS_HOME/log/hostname/racg/ - The log files for the Oracle RAC high availability component
CRS_HOME/log/hostname/racg/ - The log files for the Oracle RAC high availability component
CRS_HOME/log/hostanme/alert.log – The alert.log for Clusterware issues.

Please note that the CRS_HOME is the directory in which the Oracle Clusterware software was installed and hostname is the name of the node

Regards,
Satishbabu Gunukula
http://www.oracleracexpert.com

RMAN-06900: WARNING: unable to generate V$RMAN_STATUS or V$RMAN_OUTPUT row

One of my RMAN backup generated below Warnings, here is the detailed description and solution. I hope it helps other users.

Error:

RMAN-06900: WARNING: unable to generate V$RMAN_STATUS or V$RMAN_OUTPUT row
RMAN-06901: WARNING: disabling update of the V$RMAN_STATUS and V$RMAN_OUTPUT rows
ORACLE error from target database:
ORA-00604: error occurred at recursive SQL level 3
ORA-04031: unable to allocate 4120 bytes of shared memory ("shared pool","select name,password,datats#...","Typecheck","kgghteInit")
ORA-06508: PL/SQL: could not find program unit being called: "SYS.DBMS_BACKUP_RESTORE"

Cause & Solution:

The RMAN not able to update V$RMAN_STATUS or V$RMAN_OUTPUT due to associated error message. When you receive RMAN-06900 and RMAN-06901error, you need to look for associated error messages to find the root cause of the error message.

In this case, ORA-04031 is the associated error message

ORA-04031: unable to allocate 4120 bytes of shared memory
*Cause: More shared memory is needed than was allocated in the shared pool.
*Action: If the shared pool is out of memory, either use the dbms_shared_pool package to pin large packages, reduce your use of shared memory, or increase the amount of available shared memory by increasing the value of the INIT.ORA parameters "shared_pool_reserved_size" and “shared_pool_size". If the large pool is out of memory, increase the INIT.ORA parameter "large_pool_size".

I have seen many users complaining about error ORA-19921. If you receive this error close some of existing connections or unused RMAN connections.

If your oracle version > 10.2.0.4 then you might be hitting below bugs. These bugs fixed in 10.2.0.4 and higher versions.

ORA-19921: maximum number of 64 rows exceeded - oracle Bug 465973.
ORA-19921: maximum number of 128 rows exceeded - oracle Bug 8264365

Regards,
Satishbabu Gunukula
http://www.oracleracexpert.com

Oracle Recovery Manager(RMAN) New features in 11g

Oracle Recovery Manager (RMAN) is the preferred method to backup database and it is included with Oracle software. Using RMAN you can backup the database to local/shared disk without any other agents. If you need to backup to tape then you need Media management library.

Please refer below link for configuring RMAN to backup database to a media manager.
http://docs.oracle.com/cd/B28359_01/backup.111/b28270/rcmconfb.htm#i1006526

The restore and recovery of the database is very easy using RMAN, it will figure it out where the backups are stored and which data files need to be restored and recovered.

Oracle 11g introduced many exciting new features in Recovery Manager (RMAN). Lets start using these new features and benefit from it.

• RMAN Proactive health checks :- Now you can proactively check database block corruptions using VALIDATE DATABASE
RMAN > VALIDATE DATABASE;

You can also run the proactive check on a datafile, tablespace or block.
RMAN > VALIDATE DATAFILE 1 BLOCK 10;
RMAN > VALIDATE TABLESAPCE users;

• Database Recovery Advisor : – The data recovery advisor automatically diagnoses the corruptions and loss of data on disk and determines the corrective repair options.

The LIST FAILURE command displays any failures with priority of CRITICAL or HIGH in order of importance
RMAN> LIST FAILURE

The ADVISE FAILURE command provides repair advice for failures listed by LIST FAILURE
RMAN > ADVICE FAILURE

The REPAIR FAILURE command applies the repair scripts advised by the ADVISE FAILURE command.
RMAN>REPAIR FAILURE

You can also use PREVIEW command to see the contents of the repair before proceeding to actual repair
RMAN>REPAIR FAILURE PREVIEW

• Faster Backup Compression: The ZLIB compression is faster than original B2ZIP compression with less CPU resources

Use below command for ZLIB compression
RMAN> configure compression algorithm 'ZLIB' ;

Use below command to change compression to BZIP2
RMAN> configure compression algorithm 'bzip2';

• More backup Comcodession Choices (11gR2 only): Now we have different types of comcodession levels, i.e. LOW, MEDIUM and HIGHT and with CPU resource usage from least to highest.

Here is the example
RMAN> configure comcodession algorithm ‘medium’;

• RMAN Backups to Cloud: Now you can set your RMAN backup destination to Cloud. Amazon provides cloud computing service, but you need to use specially developed media management library.

• RMAN UNDO bypass: The RMAN backup command no longer backs up the UNDO data that is not needed for recovery. Prior to Oracle 11g, all UNDO transactions that were already committed also backed up. This backup undo optimization minimizes the backup time and storage.

• Duplicate Database from Backup (11gR2 only): Prior 11g, In order to duplicate a database using DUPLICATE TARGET DATABASE command you need connect to target database. In case if target database is down you cannot able to duplicate. From 11g, you can able to duplicate the database without connecting to target database, make sure that you have backups available on duplicate site.

Here is the example for duplicate database
connect auxiliary sys/xxxx@dup
connect catalog rman/xxx@rmancat
duplicate database 'PRD' to DUP' until time "to_date('01/01/11 10:00:00','mm/dd/yy hh24:mi:ss')"
db_file_name_convert =( “/dbs1/oradata/PRD","/dbs2/oradata/DUP")
backup location '/rman_backup' ;

Duplicate Database options
-NOREDO – Using this option no archivelogs will be applied
-UNDO_TABLESPACE – You must specify the UNDO tablesapce when you are not connected to target database.

• Set NEWNAME Flexibility while restoring Database (11gR2 only): When you are restoring database on different server with different file system structure you need to change data file path using “SET NEWNAME” command. If you have hundreds of datafiles then you need to change the path for all data files.

run
{
set newname for datafile 1 to ‘/dbs1/oradata/system01.dbf’;
set newname for datafile 2 to ‘/dbs2/oradata/users01.dbf’;
---- (Some detail removed for brevity) ----
restore database;
}

From 11g, you have flexibility using a single “SET NEWNAME” clause for all datafiles in a tablespace.

run
{
set newname for tablespace DATA to '/dbs1/oradata/data%b.dbf';
set newname for tablespace INDEX to '/dbs2/oradata/index%b.dbf';
---- (Some detail removed for brevity) ----
Restore database;
}

You can also set the path for entire database using a single command.
run
{
set newname for database to '/dbs1/oradata/%b';
restore database;
}

• TO DESTINATION Clause in BACKUP command (11gR2 only): You can specify a destination location for RMAN backups using “TO DESTINATION” clause .

RMAN> backup tablespace users to destination '/backup/rman;
You can also use this clause in ALLOCATE CHANNEL command
RMAN> run {
2> allocate channel c1 type disk to destination '/backup/rman';
backup database;
3> }

• Parallel backup of Same datafile: Now you can break the large datafiles into small sections and it reduces the backup time of large datafiles

Here is the example
run {
allocate channel c1 type disk format '/rman_backup1/%U';
allocate channel c2 type disk format '/rman_backup2/%U';
backup section size 100m
datafile 10;
}

• Automatic Block Repair (11g R2 Only): This feature automatically repairs the blocks on primary from blocks on Physical Standby. RECOVER BLOCK enhanced to repair the blocks on Primary as soon as it detects the corruption from standby when available.

• Transported Tablespace Enhancement: Prior 11g, the transported tablespace must be in read-only mode. But from 11g onwards you can transport the tablespace both in read-only and read-write mode.

• Virtual Private Catalog: The virtual catalogs are created within in same RMAN Catalog. Prior to 11g, we have only one catalog and catalog owner can able to see repository information for all databases. From 11g, you can separate the catalog database access across the departments/groups (for ex, Manufacturing, IT…etc.) using virtual catalogs feature to maximize the security.

Create a separate user for IT department to grant catalog access
SQL> create user it_user identified by xxxxx quota unlimited on users;
SQL> grant recovery_catalog_owner to it_user;

Grant catalog access on “itdb1” to virtual catalog owner “it_user”
$ rman target=/ rcvcat rman/rman@rmancat
RMAN> grant catalog for database itdb1 to it_user;

Now connect using the virtual catalog owner “it_user” and create virtual catalog
$ rman target=/ rcvcat ituser/xxxx@rmancat
RMAN> create virtual catalog;

• Merging RMAN Catalogs: If you have operations across different regions and maintaining different catalogs for each state and want to merge then you can use merge catalog feature without re-registering the databases in new catalog.
In below example we are importing rmancat2 into rmancat1
$ rman target=/ rcvcat rman/rman@rmancat1
RMAN> import catalog rman/rman@rmancat2;

• GUI interface for RMAN: Now you can use RMAN functionality through a GUI iFrom Enterprise manager

•Archive log Deletion Policy Enhancements: The archive log deletion policy has been extended in 11g for greater flexibility and protection for Dataguard environments.

Oracle 10g Syntax.
CONFIGURE ARCHIVELOG DELETION POLICY {CLEAR | TO {APPLIED ON STANDBY | NONE}}

Oracle 11g Syntax.
ARCHIVELOG DELETION POLICY {CLEAR | TO {APPLIED ON [ALL] STANDBY |BACKED UP integer TIMES TO DEVICE TYPE deviceSpecifier |NONE | SHIPPED TO [ALL] STANDBY}[ {APPLIED ON [ALL] STANDBY | BACKED UP integer TIMES TO DEVICE TYPE deviceSpecifier |NONE | SHIPPED TO [ALL] STANDBY}]...}

Regards
Satishbabu Gunukula
http://www.oracleracexpert.com

RMAN Incrementally updated backups

This feature enables you to create an image of data file and roll forward the image copy backups with regularly create incremental backups of the database. The image copy is updated with all changes up to the SCN of incremental backup that was taken.

This feature reduces the restoration time especially when Recovery time Objective (RTO) is set for mission critical databases. In case of any recovery required you never have more than one day of redo logs to apply.

For example: - Run this script on a regular basis to implement an updated backup strategy
RUN {
ALLOCATE CHANNEL C1 DEVICE TYPE DISK FORMAT ‘/oradata/rman/backup.%u’;
RECOVER COPY OF DATABASE WITH TAG 'incr_backup_update';
BACKUP INCREMENTAL LEVEL 1 FOR RECOVER OF COPY WITH TAG 'incr_backup_update' ATABASE;
}

• The BACKUP INCREMENTAL LEVEL 1... FOR RECOVER OF COPY WITH TAG... command creates a level 0 backup of the datafile, if there is no level 0 backup. In all subsequent runts, It produce’s level1 incremental backups of data file.

• The RECOVER COPY OF DATABASE WITH TAG... command causes RMAN to apply any incremental level 1 backups to a set of datafile copies with the same tag and roll forwards the image copies.

Incremental Merge backup’s strategy is the fast and easy way to restore entire database.

If you are backing up your database using Incremental merge backup then you can recovery the entire database very fast and easily by using SWITCH command. This command will point the controlfile to the backup of the data files that are present on the disk.

Let’s simulate a database failure by removing all or few data files. Make sure your database has been successfully backed up before you remove the files.

Mount the database and use SWITCH DATABASE TO COPY command to point the database to the backups available on disk.

SYS@DBTEST>startup mount
SYS@DBTEST>switch database to copy;
SYS@DBTEST>Recovery the database;
SYS@DBTEST>Alter database open;

Please note that we didn’t restored backup from backup disk or media, we have pointed the database to backup.

Regards
Satishbabu Gunukula
http://www.oracleracexpert.com