• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 952
  • Last Modified:

How to Scale Oracle (OPS)

I have two machines running on Win2K with Oracle 8.1.7 I want to do such a arrangments that my two databases running on both machines seems one to me. I am most concerned with space. I want to use one of the machine with 200GB free space.

Is it possible to create one table space in which the harddisks of another machine be refered?

Would I need to install Oracle Parallel Server to connect two machines. This way I can loadbalance my long running queries.

Or is there any other solution possible.

Please reply soon

  • 3
  • 3
  • 3
  • +4
2 Solutions
You can use DB LINKS to access the tables on the other machine.

About OPSO read the following:

What is the Oracle Parallel Server Option?
A normal Oracle installation consists of a single Oracle instance that accesses a database on the same computer system. With the parallel server option (PSO), multiple instances on different computer systems (nodes) can access the same database files simultaneously.
The OPS option is mainly used for high availability reasons. The parallel server can provide improved performance by using CPU resources across multiple computer systems. However, performance can be worse if your data is not partitioned correctly.

With Oracle Parallel Server installations, multiple instances mount the same database files. Communication between instances is managed by the Distributed Lock Manager (DLM). Note that the DLM is hardware and operating system dependent.

To address the possibility of two or more instances attempting to modify the same information simultaneously, Oracle uses up to ten additional background processes, named LCK0 through LCK9, to lock a resource in use by an instance.

The PSO is normally used on DEC VAX Clusters, UNIX Clusters, IBM's MVS SYSPLEX etc.

What are the benefits for running in Parallel Server mode?
Running Oracle in parallel mode can benefit you in the following ways:
High availability and limited failover support from Oracle8.
Speedup (increased transaction response time) - can be achieved for DSS applications.
Scaleup (increased transaction volume) - can be achieved for OLTP applications.
What are the benefits for running in Parallel Server mode?
Running Oracle in parallel mode can benefit you in the following ways:
High availability and limited failover support from Oracle8.
Speedup (increased transaction response time) - can be achieved for DSS applications.
Scaleup (increased transaction volume) - can be achieved for OLTP applications.
Are all applications suited for the Parallel Server?
Applications that can be partitioned based on function or data are perfect candidates for the Oracle Parallel Server.
Applications with 'hot' rows (the same row being accessed by processes on different nodes) will not work well. This is because data will constantly be moved (pinged) from one Oracle Instance to another.

Does the Parallel Server option require any special hardware?
Yes, the Parallel Server Option requires a server interconnect (might be as simple as a fast network connection) and a shared disk subsystem. These components are provided by your hardware vendor. Example systems that can be used for Parallel Server:
Windows NT Clusters
Unix Clusters like SUN PDB (Parallel DB).
How does one set up an Oracle Parallel Server?
Shut Down your Database:

Enable the Parallel Server Option. On Unix this is done by relinking the Oracle software.

Make the software available on all computer systems on which the Parallel Server will run on. This can be done by copying the software to all systems or to a shared disk.

Each instance requires its own set of Redo Log Files. Create additional log files:
    SVRMGR>     GROUP G4 ('RAW_FILE1') SIZE 500k,
    SVRMGR>     GROUP G5 ('RAW_FILE2') SIZE 500k,
    SVRMGR>     GROUP G6 ('RAW_FILE3') SIZE 500k;

Each instance requires its own set of rollback segments. Add Rollback Segments for New Nodes:

Edit the INIT.ORA files and number the instances 1, 2,...:
    CLUSTER_DATABASE = TRUE         (PARALLEL_SERVER = TRUE prior to Oracle9i).  
    THREAD = 1
    ROLLBACK_SEGMENTS = (r01, r02, r03, r04)
    # Include %T for the thread in the LOG_ARCHIVE_FORMAT string.
    # Set LM_PROCS to the number of nodes * PROCESSES
    # etc....

Create the dictionary views needed for the PSO:

On all the computer systems, startup the instances:

How does one start an Oracle Parallel Server?
There is no difference between the way you start a normal and a parallel server database, except that a parallel server database needs to be started from multiple nodes. The PARALLEL_SERVER=TRUE parameter needs to be set before a database can be started in parallel mode.
Before Oracle 8.0, use the following command sequence from each node:

        SVRMGR> connect INTERNAL
        SVRMGR> set retries 5
        SVRMGR> startup parallel retry      .. or SVRMGR> startup shared

You can also use the SET INSTANCE instanceN command to switch between instances (if defined in TNSNAMES.ORA).
How can I test if a database is running in shared (parallel) mode?
          if dbms_utility.is_parallel_server then
             dbms_output.put_line('Running in SHARED/PARALLEL mode.');
             dbms_output.put_line('Running in EXCLUSIVE mode.');
          end if;

How can I keep track of active instances?
You can keep track of active parallel server instances by executing one of the following queries:

I've never configured OPS on Win2K, but as far as I know, you need Windows 2000 Advanced Server installed with Cluster services on both machines and the SCSI storage devices must be on a shared bus (multi-initiator scsi or scsi over fibre channel).

u can map network drive to your one named as e.g. X:
u can store datafiles on C: as well as on X:
Get expert help—faster!

Need expert help—fast? Use the Help Bell for personalized assistance getting answers to your important questions.

Too much information for OPS. I will try simpler answer ...
Oracle Parallel Server - is a multi-node database with single Instance. Meaning, you have CompA, CompB, CompC and you have 3 different instances running on each server, all these server need to be connected to a shared disk. ( Note: Oracle is little different than DB2/Informix, they uses shared nothing). Now using OPS, CompA, CompB, CompC will look like one database to you, you can manage them as one single database.

If CompA, goes down, other Nodes will take over without any database downtime. Thats the beauty of OPS. Your database will never fail as long as you have least one node running. User will not notice even if any node failure.

With Oracle 9i, Oracle Parallel Server renamed as Real Application Cluster ( RAC). RAC is way better than OPS. OPS has so many problem with DML locks.

I am not sure what you want to achive.
If you simply want to share the disk, so you can put datafiles on the other server - yeah, you can map the disk and put your datafiles.

Mind you risk is greater. You will hit network bottleneck, and possibly a filesystem corrption.
I suggest: these days disk array is very cheap. Why not consider buying a disk array, and share them with both servers.
Hope you have enough time to read and grasp all these options. I hope it makes your life easy now!

Enayet, what i was trying to point out in my comment was that I think you can't use OPS using a network mapped disk. Everything I've read says you need a raw device (no file system).

Though I've never tried it...

In fact Mirza seems to try to use the disk resources of many computers= The answer about OPSO is comlexq because the topic ... is complex and before jumping in the sea one have to know how deep it is. Nevertheles I think that simple but good designed DBLINKs will do the job.
MirzaAmerBaigAuthor Commented:
Yes schwertner is right! My major objective is to utilize the disk space, but on the other hand if I have server then why not use processing power of it. That's why I want to use OPS. Now till now I came to know that I have to configure Cluster on two Win2k machines by installing Directory services then I would be able to install OPS. OPS is a part of Oracle enterprise edition. First I need to run two machines on Cluster.

I am looking for neat and claen process to do all. Till now didn't get it. Anyway help is great.



The best place to find out more information on configuring OPS on Win2k would be the Oracle Documentation. You can visit the otn.oracle.com and access the documentation. If you are stuck with some step, then you can post it here.

Step-By-Step Installation of RAC on Windows NT/2000

This document will provide the reader with step-by-step instructions on how to install a cluster, install Oracle Real Application Clusters (RAC) and start a cluster database on a Windows NT or 2000 cluster.  For additional explanation or information on any of these steps, please see the references listed at the end of this document.

1. Configuring the Cluster Hardware

1.1 Minimal Hardware List / System Requirements

Certified cluster configurations are provided by clicking here Oracle9i RAC Certification for Intel Platforms, which is the following URL: http://www.oracle.com/ip/deploy/database/features/ops/certification/index.html?content2.html.  Note that there are different configurations for Windows NT and 2000.  Please consult this listing for specific Hardware/Software/Variance information provided by your Cluster vendor.  In general, each node will require the following:

1.1.1.      Hardware:

o       External shared hard disks

o       Certified hardware configurations

1.1.2.      Software:

o       Certified vendor-supplied operating system dependent clusterware layer (Step 1.3)

o       Oracle Operating System Dependent (OSD) clusterware layer (Step 3.1)


1.1.3.      RAM:

o       256 MB for each instance running on that node


The above information is contained within the Oracle9i Database Installation Guide Release 1 ( for Windows.  See the section Oracle9i Database System Requirements for additional information on hardware/system sizing for other options of the RDBMS.

1.2 Installing Shared Disk Array

Follow the procedures provided by your Cluster vendor.  Verify that all nodes can view the shared partitions within the Windows NT Disk Administrator or within the Disk Manager in Windows 2000.

1.3 Installing Cluster Interconnect and Public Network Hardware

Follow the procedures provided by your Cluster vendor.  In general, you will setup the following Hostname and IP information before running the Cluster setup:

1.3.1.      Setup the External and Internal Network Interface Cards (NIC):

o       Within the Network settings of Windows, create at least two entries for the NICs you have installed.  

o       When assigning the Bindings of the NICs within the Windows Networking Properties, ensure that the Public IP is listed at the top for all settings.  The Private NIC(s) should be listed below the public NIC settings.  You can verify this at the command prompt by running the command ipconfig /all to verify that the public IP address is listed first.  


1.3.2.      Resolution of External and Internal Hostnames:

o       Ensure that the External or Public Hostnames are defined in your Directory Network Services (DNS) and that the correct IP addresses resolve for all nodes in the cluster.

o       Ensure that the Internal or Private Hostnames are defined in the HOST file on all nodes of the cluster.  This file is located in the WINDOWS_HOME\System32\drivers\etc directory.

o       Test your cluster configuration by pinging the hostnames from each node and check for proper names resolution.  


1.4 Checking Access to Other Nodes Within Windows

To install and perform administrative tasks, Oracle Corporation recommends using the same username and password on each node in a cluster, or a domain username with administrative privileges on all local nodes.  All nodes must be in the same domain.  

Ensure that each node has administrative access to all these directories within the Windows environment by running the following at the command prompt:  

NET USE \\host_name\C$

where host_name is the public network name for the other nodes.  If you plan to install the ORACLE_HOME onto another drive location than C, check that administrative share as well.  

For example, if your WIN_HOME is on the C drive and you were installing the ORACLE_HOME onto the E drive of all nodes, you would run the following from a command prompt on node 1 of a four-node cluster:

NET USE \\node2\C$

NET USE \\node3\C$

NET USE \\node4\C$

NET USE \\node2\E$

NET USE \\node3\E$

NET USE \\node4\E$

You would then repeat these commands on all nodes within the cluster.  If the following appears for each command, the privileges are correct:

The command completed successfully.  

If you receive errors, resolve these within the Windows environment before proceeding.  



2.  Configuring the Raw Partitions on the Shared Disk Array


2.1 Partitions required for the installation of a RAC database

Real Application Clusters requires that all each instance be able to access a set of unformatted devices on a shared disk subsystem.  These shared disks are also referred to as raw devices.  The Oracle instances in Real Application Clusters write data onto the raw devices to update the control file, server parameter file, each datafile, and each redo log file.  All instances in the cluster share these files.  

The Oracle Database Configuration Assistant (DBCA) will create a seed database expecting a particular configuration for a two-node cluster.  For the seed database that will be created by the DBCA, it is recommended you create the Logical Partition at least 1M larger than the desired datafile size.    The following table contains a listing of the Tablespaces, Minimum datafile size, and expected Symbolic Link name for the corresponding file:  



Tablespaces  Minimum File Size
  Expected File Name  
SYSTEM tablespace
  400 Mb
USERS tablespace
  120 Mb
TEMP tablespace
  100 Mb
UNDOTBS tablespace Thread 1
  312 Mb
UNDOTBS tablespace Thread 2
  312 Mb
CWMLITE tablespace
  100 Mb
  160 Mb
INDX tablespace
  70 Mb
TOOLS tablespace
  12 Mb
DRSYS tablespace
  90 Mb
First control file
  110 Mb
Second control file
  110 Mb
Redo Log 1 Thread 1
  120 Mb
Redo Log 2 Thread 1
  120 Mb
Redo Log 1 Thread 2
  120 Mb
Redo Log 2 Thread 2
  120 Mb
  5 Mb
  100 Mb

Notes on Table:  Automatic Undo Management requires an undo tablespace per instance; therefore, you would require a minimum of 2 tablespaces as described above.  In the sample names listed in the table, the string db_name should be replaced with the actual database name.  The raw partition files within Windows are identified with the \\.\ notation and usually do not have an extension in the filename.  

The basic procedure to create the shared partitions is to first create the Extended Partitions for the hard disks in the array, then create as many Logical Partitions as required for the database files.  Only one extended partition can be created for each disk.  Oracle Corporation recommends creating the extended partition on an unpartitioned disk and using the entire disk for the extended partition.

Please refer to the section Planning Your Logical Drives Configuration in Appendix B of the Oracle9i Database Installation Guide Release 1 ( for Windows for more guidelines on this.  

Note:  Creating a large number of logical partitions may cause a significant increase in the time needed to reboot and start the disk administration tools.  Oracle Corporation recommends you do not create more than 120 logical drives in an extended partition.
2.2 Creating an Extended Partition and Logical Drives on Windows 2000
o       Log in as member of the Local Administrators Group.

o       Choose Settings > Control Panel.  Double click on the Administrative Tools > Computer Management.  In the console tree, select Disk Management.  

o       To create the Extended Partitions, right-click the unallocated region of a basic disk, and choose Create Partition.  

o       In the Create Partition wizard, choose Next > Extended Partition.  

o       Choose Extended Partition. Choose Next.

o       Choose the maximum amount of space and click Next.

o       A summary screed will come up.  Choose Finish.

o       Repeat until all Extended Partitions are complete.

o       To create the Logical Partitions, right-click again on the Extended Partition and choose Create Logical Drive.

o       Follow the instructions in the wizard, choosing the appropriate size for your desired partition, choosing to not assign any drive letters, and choosing no format.  

o       Repeat until all Logical Drives are complete.

Note:  If the Disk Management window is open during any disk management modifications, such as creating symbolic links or adding logical partitions, you need to close and open the window to view any changes you applied.

2.3 Creating an Extended Partitions and Logical Drives on Windows NT
o       Log in as member of the Local Administrators Group.

o       Choose Start > Programs > Administrative Tools > Disk Administrator for Windows NT.  The Disk Administrator window appears, showing all the disks in your shared array.  Note that the lines display diagonally from top right to bottom left, indicating unpartitioned devices.

o       To create the Extended Partitions, right-click the unpartitioned disk and choose the Disk Administrator Create Extended option in the menu.

o       Select Create Extended.  The Disk Administrator displays the maximum sizes for the extended partition.  Accept and choose OK.

o       Repeat until all Extended Partitions are complete.

o       To create the Logical Partitions, select an area of free space in the extended partition and right-click.  Choose Partition > Create.

o       Enter the desired size for the partition and click OK.  

o       Right-click on the Logical Drive and choose Tools > Assign Drive Letter.  Click the Do Not Assign Letter option.

o       Repeat until all Logical Drives are complete.

Note:  Changes are not saved until you choose Commit Changes Now or exit Disk Administrator.

2.4 Verify changes are visible on all nodes in the cluster.

o       Open the Disk Administrator or Disk Management tool on all nodes.  Ensure that all partitions can be seen and that there are no drive letters assigned to any of these partitions.  

o       If there are any anomalies, correct within the Windows environment before proceeding.



3.  Configuring the Cluster using the Oracle OSD Software


Note:  The Oracle Cluster Setup Wizard installs Oracle9i OSD clusterware and raw device management utilities.  If you do not use Cluster Setup Wizard, then the Oracle Object Link Manager and the raw device management utilities do not get installed until Oracle Universal Installer is run.  Before the Oracle Universal Installer is invoked, it is necessary to configure and name the raw devices using these utilities.  Therefore, this document will outline the Oracle Cluster Setup Wizard procedure.  

If you intend to use vendor operating system dependent clusterware instead of Oracle9i operating system dependent clusterware, then you can temporarily install the raw device management utilities located in the Disk1\preinstall_rac\olm directory.  See the readme.txt for a description of each utility.  See the section "Installing the Raw Devices Management Utilities Manually" in the Oracle9i Database Installation Guide Release 1 ( for Windows for installation instructions.  

3.1 Perform a Final Cluster Check

At the command prompt, run the clustercheck.exe program located in the Disk1\preinstall_rac\clustercheck directory.  This tool will prompt for the public and private hostnames and have you verify the IP address resolution.  If that passes, then it will perform a check the health of the shared disk array and other environment variables and permissions necessary for proper cluster installation and operation.  It will create a subdirectory called opsm in the temporary directory specified by your environment settings (WIN_DRIVE:\Documents and Settings\Administrator\Local Settings\Temp by default) and log file called OraInfoCoord.log.  This log will contain any errors encountered in the check.  You should see the following at the bottom of the log file and within the command prompt window when you run the clustercheck.exe program:


You must correct any errors that occur before proceeding with the Oracle OSD Software installation.  Please contact your Cluster Hardware Vendor if you need assistance.  

3.2 Installing the Oracle OSD Software

The Oracle Cluster Setup Wizard performs the following tasks on all nodes:

Installs and starts Oracle9i operating system dependent clusterware
Installs Object Link Manager and starts the Oracle Object Service on all nodes.  This tool creates persistent symbolic links to the logical drives required for an Oracle Real Application Clusters database.  The service updates all nodes when symbolic links are modified, and is set to Automatic startup so that it starts whenever you reboot the node.
Preserves existing symbolic link information created by previous invocations of Oracle Object Link Manager
Installs other disk management tools on all nodes
Adds a node to an existing cluster
To install the OSD software, navigate to the \preinstall_rac\clustersetup directory on the Oracle9i Server Installation CD:

o       Launch the Oracle Cluster Setup Wizard by double-clicking on clustersetup.exe.  Then choose Next.

o       The first time the Wizard is run, the only option will be to Create a cluster.  Choose Next.

o       The Disk Configuration screen appears.  Click on the Create Oracle Symbolic Links button.

o       The Oracle Object Link Manager window appears:

1.      From the Symbolic Link column, select an empty row.  The cursor starts blinking.

2.      Enter the desired datafile link name and hit Enter to save.

3.      Repeat steps 1 and 2 to assign all symbolic link names required (as outlined in the table in section 2.1).

4.      Click on the Apply button to commit the changes.

5.      When the progress bar at the bottom of the screen stops moving, choose Close.

From the Disk Configuration screen, assign a Voting disk, labeled as srvcfg, by highlighting the corresponding row.  Choose Next.
The VIA Detection screen appears stating whether Virtual Interface Architecture (VIA) hardware was detected.  Choose yes or no depending on your configuration.  Please contact your cluster hardware vendor if you are unsure. Click Next.
The Network Selection screen appears.  Choose Use private network for interconnect.  Click Next.
The Network Configuration screen appears.  Enter the public hostnames for all nodes.  The private hostnames will be automatically entered as public_name.san.  Change as appropriate for your cluster configuration.  Click Next.
The Install Location screen appears.  It will default to the WIN_HOME\system32\osd9i directory.   Accept the default and click Finish.
The Cluster Setup window will appear.  This will show the progress with installing the cluster files and creating the cluster services on all nodes.  If no errors occur, the Oracle Cluster Setup Wizard application will complete and close automatically.  

3.3 Check the Cluster Install on All Nodes

Verify that all nodes have the following services running and set to Automatic startup within Control Panel > Services:

o       OracleObjectService

o       OracleCMService

Verify that all nodes have an osd9i subdirectory in the WIN_HOME\system32 directory.  Also verify all node registries contain the following key:  HKEY_LOCAL_MACHINE\Software\Oracle\osd9i.  



4. Using the Oracle Universal Installer for Real Application Clusters

The Oracle Universal Installer will install the Oracle9i Server Enterprise Edition and the Real Application Clusters software.  Oracle9i Server Enterprise Edition for Windows is supplied in 3 CD-ROM disks.  To install the Oracle Software, perform the following:

The Oracle Universal Installer should launch by Autostart.  If not, browse the CD within Windows Explorer to the setup.exe file and double-click it.    
At the OUI Welcome screen, click Next.
A prompt will appear for the Source path where the products.jar file is located.  If this is the first time that OUI has been run on this system, this should default to the CD.  If not, browse to the stage directory on the CD and highlight the file.  For the Destination section, enter a name and location for your Oracle9i Home.  This is the base directory into which OUI will install files. Click Next and a bar at the top of the window will show the progress of loading the products list. When it reaches 100%, it will proceed to the next screen.
The Available Products screen appears.  Select the Oracle9i Database, and then click Next.
The Installation Type screen appears.  Choose the Enterprise Edition.  The selection on this screen refers to the installation operation, not the database configuration.  Click Next.
The Database Configuration screen appears.  Choose to perform a Software Only installation.  Click Next.
The Cluster Nodes Selection screen appears.  Select the other nodes on to which the Oracle RDBMS software will be installed.  It is not necessary to select the node on which the OUI is currently running.  If all nodes are present, click Next.
NOTE:  If not all nodes in the cluster are listed, there is something wrong with your cluster configuration  you will have to go back and troubleshoot your cluster install.  You can perform clusterware diagnostics by executing the ORACLE_HOME\bin\lsnodes -v command and analyzing its output.  Refer to your vendor's clusterware documentation if the output indicates that your clusterware is not properly installed.  Resolve the problem and then restart the Oracle Cluster Wizard.

The Summary screen will be presented.  Confirm the RAC database software will be installed (it will be listed as Oracle9i Real Applications Cluster) and then click Install.  
The OUI will install the Oracle9i software on to the local node, and then copy this information to the other nodes selected.  This will take some time.  During the installation process, the OUI does not display all the messages indicating components are being installed on other nodes, so the installation may appear to be hung.  In this case, I/O activity may be the only indication that the process is continuing.  If necessary, check each nodes activity using Task Manager.

Verify that the install completed by checking for an OracleGSDService on all nodes.  If the service did not get created and started, begin by typing the following at the command line on each node:

      C:\> gsdservice start

This service must be running before starting the Oracle Database Configuration Assistant (DBCA).  


5. Create a RAC Database using the Oracle Database Configuration Assistant

The Oracle Database Configuration Assistant (DBCA) will create a database for you.  The DBCA creates your database using the optimal flexible architecture (OFA).  This means the DBCA creates your database files, including the default server parameter file, using standard file naming and file placement practices.  The primary phases of DBCA processing are:

Verify that you correctly configured the shared disks for each tablespace
Create the database
Configure the Oracle network services
Start the database instances and listeners
Oracle Corporation recommends that you use the DBCA to create your database.  This is because the DBCA preconfigured databases optimize your environment to take advantage of Oracle9i features such as the server parameter file and automatic undo management.  The DBCA also enables you to define arbitrary tablespaces as part of the database creation process.  So even if you have datafile requirements that differ from those offered in one of the DBCA templates, use the DBCA.  You can also execute user-specified scripts as part of the database creation process.  The DBCA and the Oracle Net Configuration Assistant also accurately configure your Real Application Clusters environment for various Oracle high availability features and cluster administration tools.

Manually launch the Oracle Database Creation Assistant by navigating to the Oracle Program group, then to Configuration and Migration Tools.  
The Welcome Page displays.  In a moment, a selection screen appears to create a Cluster or Single Instance Database.  Choose Oracle Cluster Database option and select Next.
The Operations page is displayed. Choose the option Create a Database and click Next.
The Node Selection page appears.  Select the nodes that you want to configure as part of the RAC database and click Next.  Please see this Note if not all nodes are listed.
The Database Templates page is displayed.  The templates other than New Database include preconfigured datafiles for file systems.  Choose New Database and then click Next.
DBCA now displays the Database Identification page.  Enter the Global Database Name and Oracle System Identifier (SID).  The Global Database Name is typically of the form name.domain, for example mydb.us.oracle.com, while the SID is used to uniquely identify an instance (DBCA should insert a suggested SID, equivalent to name1 where name was entered in the Database Name field).  In the RAC case, the SID specified will be used as a prefix for the instance number.  For example, MYDB would become MYDB1 and MYDB2 for instances 1 and 2, respectively.
The Database Options page is displayed.  Select the options you wish to configure.  The Additional database Configurations button displays the option to install Java and interMedia database features.  Check all options you wish and then choose Next.  Note:  If you did not choose New Database from the Database Template page, you will not see this screen.  
Select the connection options desired from the Database Connection Options page.  Note:  If you did not choose New Database from the Database Template page, you will not see this screen.  Click Next.
DBCA now displays the Initialization Parameters page.  This page comprises a number of Tab fields which you navigate through by clicking on the tabs:
Modify the Memory settings if desired.
Change the Archivelog mode as necessary.  In general, it is recommended you create your database in Noarchivelog mode, and then after the database is created, alter the database after performing a complete backup.
DB Sizing will specify your db_block_size, sort_area_size and database character set parameters.  
Under the File Locations tab, the option Create persistent initialization parameter file is selected by default.  The raw device name for the location of the server parameter file (spfile) must be entered.  The button File Location Variables displays variable information.  Click OK.
The button All Initialization Parameters displays the Initialization Parameters dialog box.  This box presents values for all initialization parameters and indicates whether they are to be included in the spfile to be created through the check box, included (Y/N).  Instance specific parameters have an instance value in the instance column.  Complete entries in the All Initialization Parameters page and select Close.  
Note:  There are a few exceptions to what can be altered via this screen. Ensure all entries in the Initialization Parameters page are complete and select Next.

DBCA now displays the Database Storage Window.  This page allows you to enter file names for each tablespace in your database.  The file names are displayed in the Datafiles folder, but are entered by selecting the Tablespaces icon, and then selecting the tablespace object from the expanded tree.  Any names displayed here can be changed.  These should be listed as you have entered in Step 3.  Complete the database storage information and click Next.
The Database Creation Options page is displayed.  Ensure that the option Create Database is checked and click Finish.  Check the Create template and save as a script boxes if desired.
The DBCA Summary window is displayed.  Review this information and then click OK. Once the Summary screen is closed using the OK option, DBCA begins to create the database according to the values specified.
Note:  The database creation can take a while, and the progress may seem slow or hung, especially during the section where the DBCA is running scripts.  You can check the progress by checking Task Manager and seeing the CPU activity, or by checking the alert log for redo log switching.

A new database now exists. It can be accessed via Oracle SQLPlusWorksheet or other applications designed to work with an Oracle RAC database




6.  Using SRVCTL for the Administration and Maintenance of a RAC database


Once your RAC database is created, you can use the Server Control (SRVCTL) utility to assist in administration and maintenance tasks.   The Global Services Daemon (GSD) receives requests from SRVCTL to execute administrative jobs, such as startup or shutdown.  The task is executed locally on all nodes, and the results are sent back to SRVCTL.  SRVCTL also serves as a single point of control between the Oracle Intelligent Agent and the nodes in the cluster.  

To see the online command syntax and options for each SRVCTL command, enter:  

srvctl command option -h

Where command option is one of the valid options such as start, stop, or status.  

The following are some example of tasks you can perform with this utility.

o       srvctl start
Use this command to start all instances or a subset of instances in your Real Application Clusters database.  For example, to start all the instances use the syntax:

  srvctl start -p db_name
Or you can start specific instances using the syntax:

start -p db_name -i instance_name
This syntax starts the specific instance that you name.  Using srvctl start also starts all listeners associated with an instance.

Use the -s option (lower-case s) as in the following syntax where stage is either inst or lsnr:

  srvctl start -p db_name -s stage
Use this syntax to separately start instances and listeners.

o       srvctl stop
Use this command to stop all instances or a subset of instances in your Real Application Clusters database.  For example, to stop all instances use the syntax:

  srvctl stop -p db_name
Or you can stop specific instances using:

stop -p db_name -i instance_name
Using srvctl stop also stops all listeners associated with an instance.

Use the -s option (lower-case s) as in the following syntax where stage is either inst or lsnr:

  srvctl stop -p db_name -s stage
Use this syntax to separately stop instances and listeners.

o       srvctl status
Use the srvctl status command to determine what instances are running. For example, use the output from the following syntax to identify which instances are running:

  srvctl status -p db_name
Use the -s option (lower-case s) as in the following syntax where stage is either inst or lsnr:

  srvctl status -p db_name -s stage
Use this syntax to obtain the status of an instance or a listener.

o       srvctl config
Use the srvctl config command to identify the existing Real Application Clusters databases. You can use two syntaxes for srvctl config. For example, the following syntax lists all the Real Application Clusters databases in your environment:

  srvctl config
The following syntax lists the instances for the Real Application Clusters database name that you provide:

  srvctl config -p db_name
The Oracle Enterprise Manager auto-discovery process also uses output from this command to discover the configurations for databases in your Real Application Clusters.

o       srvctl get env
Use the srvctl get env command to obtain environment information for either a specific instance or for an entire Real Application Clusters database. For example, the output from the following syntax displays environment information for the entire Real Application Clusters database identified by the name you provide:

  srvctl get env -p db_name
The following syntax displays environment information for a specific instance:

get env -p db_name -i instance_name





7.   References

Oracle9i Database Installation Guide Release 1 ( for Windows, Part Number A90162-01
Oracle9i Real Application Clusters Installation and Configuration Release 1 (9.0.1)
Oracle9i Real Application Clusters Concepts
Oracle9i Real Application Clusters Administration
Oracle9i Real Application Clusters Deployment and Performance
Oracle9i Release Notes
Oracle9i for Windows 2000 Tips and Techniques:  Best Practices from Oracle Experts  Oracle Press
My humble suggestion is: don't even bother with RAC, its fairly complex! unless you really want to! There are more to consider before using it ....

Shared disk array is your best choice. Or DBLINK as some suggested. Maping network drive also an option for you ...

Good luck!
A database link is not a bad idea. OPS and RAC are a bit painful to get used to and it sounds as if you don't have the required hardware anyway.

Consider partitioning your application instead. That is, try to figure out if there are ways to divide up the work between the machines. For example, if you have a data warehouse, you might consider keeping the schemas on the larger machine and using it for aggregation, and keeping the summary data on the smaller machine for users to query against.

MirzaAmerBaigAuthor Commented:
Although many has responded with sugestions, but still I can't conclude what steps should I take to solve the OPS and cluster issue.

I will wait for few days then may close the question.

Amer -
OPS (RAC) could be a good idea if you want to effectively use two servers and their sources.

The reason I discourage RAC, it has so much pain inflicted for DBAs. Lock Manager for OPS is a total pain! I have experienced it:( But I hear, 9i RAC ( OPS for 8i) is much better. Oracle introduced few new technology to reduce lock and latch contention. SO its much better.

You can use 9i RAC (OPS for 8i), if you want.

Hope this help!
No comment has been added lately, so it's time to clean up this TA.
I will leave a recommendation in the Cleanup topic area that this question is:
Split between schwertner & enayet
Decision based on sheer amount of relevant information provided and PAQ value.
Please leave any comments here within the next seven days.
EE Cleanup Volunteer
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Cloud Class® Course: Microsoft Azure 2017

Azure has a changed a lot since it was originally introduce by adding new services and features. Do you know everything you need to about Azure? This course will teach you about the Azure App Service, monitoring and application insights, DevOps, and Team Services.

  • 3
  • 3
  • 3
  • +4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now