1
Introduction to the Installation and Configuration Process

This chapter provides an overview of the Oracle Parallel Server installation process by describing the components required for configuring and deploying Oracle Parallel Server. This information also helps you prepare and plan your Oracle Parallel Server installation and configuration process in an Oracle8i environment.

This chapter includes the following topics:

• Introduction to Oracle Parallel Server

• Oracle Parallel Server Software Components

• Installation Overview

• Database Configuration Overview

• Selecting a Database Creation Method

Introduction to Oracle Parallel Server

Oracle Parallel Server is an architecture that enables multiple instances to access a shared database. Oracle Parallel Server offers the following:

• Superior scalability through distribution of workload across nodes

• High-availability through multiple nodes accessing the database; if one node fails, the database is still accessible through surviving nodes

Each Oracle Parallel Server instance, a running instance of Oracle8i software, is comprised of a System Global Area (SGA) and Oracle background processes. The SGA is a shared memory region that contains data and control information for an instance. Users can connect to any instance to access the information that resides within the shared database.

See Also: Oracle8i Parallel Server Concepts

Oracle Parallel Server instances coordinate with the following components:

Component	Description
Node	A server where an instance resides
Cluster	A set of physically interconnected nodes and a shared disk storage subsystem
Database	The set of all instances with access to the same data files. A database is limited to a set of Oracle Parallel Server instances that run on only the nodes defined within a cluster. The database name is defined by global database name, a name comprised of the database name and domain, entered during installation. The database name is unique in the cluster.

All instances share:

• The same set of data files

• The same control files to start and run the database

The data files and control files are located on raw disk devices that are shared between multiple nodes.

If one node fails, client applications (written to do so) can re-route users to another node. One of the surviving nodes automatically performs recovery by rolling back any incomplete transactions that the other node was attempting. This ensures the logical consistency of the database.

Note: Because an instance does not include data files, you can start up an instance without mounting the data files.

Each instance has its own set of redo log files. Although instances use their own redo log files, the redo log fields must exist on the raw devices. This way, each node can perform recovery on behalf of another node.

Figure 1-1 shows the relationship between two Oracle instances and the shared disks on which the data files are stored:

Figure 1-1 Oracle Parallel Server Instance and Shared Data File Architecture

Oracle Parallel Server Software Components

The following components comprise Oracle Parallel Server:

Component	Description
Oracle8i Enterprise Edition	Provides the applications and files to manage a database. All other Oracle Parallel Server components are layered on top of Oracle8i Enterprise Edition
Oracle Parallel Server	Provides the necessary Oracle Parallel Server scripts, initialization parameter files, and data files
Oracle Parallel Server Management	Provides a single point for starting, stopping, and monitoring the activity of parallel servers and parallel server instances from within Oracle Enterprise Manager Console See Also: Chapter 5 Chapter 6
Operating System Dependent layer	Consists of several software components developed by vendors. The Operating System Dependent layer maps the key OS/cluster-ware services required for proper operation of Oracle Parallel Server. See Also: "Operating System Dependent Layer" below

Operating System Dependent Layer

You must install a vendor-supplied, Oracle-certified Operating System Dependent layer before you install Oracle Parallel Server software. The Operating System Dependent layer maps the key OS/cluster-ware services required for Oracle Parallel Server. This layer consists of two primary software components developed by vendors, as described in the following table:

Component	Description
Cluster Manager (CM)	Discovers the state of the cluster and instance membership
Inter-Process Communication (IPC)	Provides reliable transfer of messages between instances on different nodes

These components provide key services required for proper operation of Oracle Parallel Server and are used by various clients, such as Integrated Distributed Lock Manager.

Figure 1-2 illustrates the Operating System Dependent components in a cluster with two nodes:

Figure 1-2 Operating System Dependent Components

Installation Overview

Oracle Parallel Server software installation requires the following components:

• Operating System Dependent layer

• Oracle8i Enterprise Edition

• Oracle Parallel Server

  
  

Similar to building blocks, the components must be installed in the following order:

1. Install the Operating System Dependent layer, as prescribed by your vendor.

2. Install Oracle8i Enterprise Edition and Oracle Parallel Server from the Oracle8i Enterprise Edition CD-ROM.

   Note: If you intend to build an Oracle Parallel Server database during installation, follow the raw device creation guidelines in Chapter 2. Raw device creation is necessary because Oracle Database Configuration Assistant, which is launched during installation, expects the devices to be established.

Database Configuration Overview

Once Oracle Parallel Server components are installed, the database can be created. Database creation involves two steps:

1. Establish raw devices for:
   • Shared data files

   • Shared control files

   • Each node's redo log files

2. Create the database using either Oracle Database Configuration Assistant application or manual methods.

Selecting a Database Creation Method

The database can be created using Oracle Database Configuration Assistant or using manual methods. The table below describes the two methods:

Creation Method	Description
Oracle Database Configuration Assistant	Oracle Database Configuration Assistant is a graphical user interface (GUI) tool that enables you to create a database for an Online Transaction Processing (OLTP), Data Warehousing, or Multi-Purpose environment. Oracle Database Configuration Assistance creates a database using the Optimal Flexible Architecture (OFA), whereby database files and administrative files, including initialization parameter files, follow standard naming and placement practices. Additionally, it creates a standard set of tablespaces. Oracle Database Configuration Assistant can be automatically started from within Oracle Universal Installer when you select to create a database or manually run as a stand-alone tool.
Manual	The database should be created manually if you already have existing scripts or require different data files and tablespaces than those created by Oracle Database Configuration Assistant.

Using Oracle Database Configuration Assistant

This section contains the following sections:

• Advantages of Using Oracle Database Configuration Assistant

• Identifying Your Database Environment

• Selecting a Database Creation Method

• Looking at a Database Created by Oracle Database Configuration Assistant

Advantages of Using Oracle Database Configuration Assistant

Here are some of the advantages of using Oracle Database Configuration Assistant:

• It uses Optimal Flexible Architecture (OFA).

• It is fast. A ready made database can be copied into place rather than going through a lengthy creation process. This database is a Multi-Purpose database, suiting many users' environments. It can be customized later, if desired.

• You do not need to spend time deciding how to set parameters. Decisions have already been made for you.

• It customizes your database for you. Even if you do not choose to copy the starter database, you can still direct Oracle Database Configuration Assistant to generate a script that creates an online transaction processing (OLTP), Data Warehousing, or Multi-Purpose database. You need answer only a few questions presented to you by Oracle Database Configuration Assistant. It automatically includes multi-threaded server (MTS) parameters if you have over a specified number of users.

Identifying Your Database Environment

Oracle Database Configuration Assistant enables you to create a database in one of the following environments. Identify the environment appropriate for your database:

Environment	Description
Online Transaction Processing (OLTP)	Databases in OLTP environments must process thousands or even millions of transactions from many concurrent (simultaneously connected) users each day. These transactions consist of reading (SELECT statements), writing (INSERT and UPDATE statements), and deleting (DELETE statements) data in database tables. Users must have quick access to the most current data. Therefore, database performance is defined in terms of throughput (speed) and availability of data.
Data Warehousing	Databases in Data Warehousing environments must process a variety of queries (typically read-only), ranging from a simple fetch of a few records to numerous complex queries that sort thousands of records from many different tables. Therefore, database performance is defined in terms of response time.
Multi-Purpose	Multi-Purpose databases support both OLTP and Data Warehousing environments.

Selecting a Database Creation Method

When you run Oracle Universal Installer and select Oracle8i Enterprise Edition, you are presented with choice of Typical or Custom installation. Each installation type enables you to create a database.

The types of database (OLTP, Data Warehousing, and Multi-Purpose) created with the Typical and Custom installation types and the amount of user input are described below. Review these selections to identify the database that best matches your database requirements and database creation expertise.

If You Perform These Steps...	Then...
Select the Typical installation type.	Oracle Database Configuration Assistant automatically starts at the end of installation and copies a preconfigured, ready-to-use Multi-Purpose database from the CD-ROM. This is referred to as a seed database or starter database. No user input is requested, making creation efficient.
Select the Custom installation type. Select Yes when prompted to create a database Oracle Database Configuration Assistant prompts you to select either of two choices: - Typical database creation type - Custom database creation type	Oracle Database Configuration Assistant role in database creation depends upon your selection: Typical database creation type You have two choices. Oracle Database Configuration Assistant role in database creation depends upon your selection. If you select: Copy existing database files from the CD Oracle Database Configuration Assistant creates the same database as during Typical installation. No user input is required. Create new database files Asks you several database environment questions before dynamically creating a database script for an OLTP database, Data Warehousing database, or Multi-Purpose database with standard initsid.ora and initdb_name.ora files. Note: If 20 or more users are specified for an OLTP database, MTS parameters are set in the initdb_name.ora file. Custom database creation type Oracle Database Configuration Assistant guides you in the creation of database fully customized to match the environment (OLTP, Data Warehousing, Multi-Purpose) and database configuration mode (dedicated server or multi-threaded server) you select. Select this option only if you are experienced with advanced database creation procedures, such as customizing: MTS parameters Data, control, and redo log file settings Tablespace sizes Database memory parameters Archiving formats and destinations Trace file destinations Character set value

Looking at a Database Created by Oracle Database Configuration Assistant

The database configured with Oracle Database Configuration Assistant contains the following features:

• Tablespaces and Data Files

• Redo Log Files

• Control Files

• Rollback Segments

• Initialization Parameter Files

• Unique Identification of Instances with the initsid.ora File

Tablespaces and Data Files

An Oracle Parallel Server database is divided into smaller logical areas of space known as tablespaces Each tablespace corresponds to one data file, stored on a raw device. The tablespace names used in an Oracle Parallel Server database contain the following types of data:

Tablespace Name	Contents
SYSTEM	Consists of the data dictionary, including definitions of tables, views, and stored procedures needed by the database. Information in this area is maintained automatically. The SYSTEM tablespace is present in all Oracle8i databases.
USER	Consists of application data. As you create and enter data into tables, you fill this space with your data.
TEMP	Temporary tables and/or indexes created during the processing of your SQL statement. You may need to expand this tablespace if you are executing a SQL statement that involves a lot of sorting, such as ANALYZE COMPUTE STATISTICS on a very large table, or the constructs GROUP BY, ORDER BY, or DISTINCT.
RBS	Consists of the rollback segments. Rollback segments are sets of extents that record data before it is modified by transactions, enabling the changes to be undone.
INDX	Stores indexes associated with the data in the USER tablespace
TOOLS	Stores tables for Oracle Enterprise Manager
DRSYS	Consists of data for Oracle8i interMedia Text

The tablespace names cannot be altered with the Typical database creation type. You can change the names of the tablespaces, except for SYSTEM, with the Custom database creation type.

Each tablespace is represented by one database file. The data file names created by the Typical database creation type vary by operating system. UNIX prompts you to set the file names. Windows NT uses the symbolic link names shown in the following table:

Tablespace	Windows NT Symbolic Link
SYSTEM	\\.\db_name_system1
USERS	\\.\db_name_users1
TEMP	\\.\db_name_temp1
RBS	\\.\db_name_rbs1
INDX	\\.\db_name_indx1
TOOLS	\\.\db_name_tools1
DRSYS	\\.\db_name_drsys

You can specify different symbolic names with the Custom database creation type

Redo Log Files

Each node is configured with two redo log files. Redo log files are used to hold a record of all changes made to data. If a database failure occurs before the data is written to disk, the changes can be obtained from the redo log files. This prevents data from ever being lost. Redo log files are stored on the raw devices.

Redo log files are used in a cyclical fashion. The first file is filled first, then the second file. The first file is then re-used and filled, the second file is re-used and filled, and so on.

The redo log files created with the Typical database creation type vary by operating system. UNIX prompts you to set the file names. Windows NT uses symbolic link names of \\.\db_namethread_number, where thread is the thread ID of the node, and number is the number (1 or 2) of the redo log file. The Custom database creation type prompts you to specify redo log file names or symbolic link names.

Control Files

The database is configured with two control files. A control file is an administrative file required to start and run the database. The control file records the physical structure of the database. For example, a control file contains the database name, and the names and locations of the database's data files and redo log files. The control files are stored on the raw devices.

The controls files created by the Typical database creation type vary by operating system. UNIX prompts you to set the file names. Windows NT uses symbolic link names of \\.\control1 and \\.\control2. The Custom database creation type prompts you to specify control file names or symbolic link names.

Rollback Segments

Each node is configured with two rollback segments. Rollback segments record the old values of data changed by each transaction (whether or not committed). Every database contains one or more rollback segments, which are portions of the database that record the actions of transactions in the event that a transaction is rolled back. Rollback segments are used to provide read consistency, to roll back transactions, and to recover the database. Note that rollback segments are database objects, not files that reside in the RBS tablespace.

The rollback segments created by the Typical or Custom database creation type have names of rbsthread_number, where thread is the thread ID of the node, and number is the number (1 or 2) of the rollback segment.

Initialization Parameter Files

Each node has an instance-specific initialization parameter file and an Oracle Parallel Server database initialization parameter file.

The Typical database creation type stores the initialization parameter file in $ORACLE_BASE/admin/db_name/pfile on UNIX and ORACLE_BASE\admin\db_name\pfile on Windows NT. The table below describes the initialization parameter files:

Initialization Parameter File	Description
initsid.ora	Each node's instance has its own initsid.ora file, where sid is the Oracle System Identifier (SID) of the instance. This file uniquely defines the instance with instance-related parameters. This file calls theinitdb_name.ora file.
initdb_name.ora	Parameters for an Oracle Parallel Server database are stored in initdb_name.ora file, where db_name is the database name. It lists the common database parameters shared from node-to-node. This file resides on each of the nodes.

The Custom database type allows you to set the location of the initdb_name.ora file.

See Also: "Understanding the Initialization Parameter Files" for further information about these files.

Unique Identification of Instances with the initsid.ora File

The initsid.ora file is influenced by a number of elements, including:

Element	Description
Oracle System Identifier (SID)	Uniquely identifies a node's instance The database name, specified by the DB_NAME parameter in the initdb_name.ora file, and unique thread ID make up each node's SID. The thread ID starts at 1 for the first instance in the cluster, and is incremented by 1 for the next instance, and so on. The SID is defined as an ORACLE_SID environment variable on UNIX platforms and in the ORACLE_SID registry entry on Windows NT.
Instance Name	Represents the name of the instance and is used to uniquely identify a specific instance when multiple instances share common service names. The instance name is the same as the SID. It is indicated by the INSTANCE_NAME parameter in the initsid.ora file.
Instance Number	Associates extents of data blocks with particular instances The instance number ensure that an instance uses the extents allocated to it for inserts and updates. The instance cannot use data blocks in another free list unless the instance is restarted with that instance number. You can use various SQL options with the INSTANCE_NUMBER initialization parameter to associate extents of data blocks with instances. The instance number is indicated by the INSTANCE_NUMBER parameter in the initsid.ora file.
Rollback Segments	Each node is configured with two rollback segments. Rollback segments record the old values of data changed by each transaction (whether or not committed). Every database contains one or more rollback segments, which are portions of the database that record the actions of transactions in the event that a transaction is rolled back. Rollback segments are used to provide read consistency, to roll back transactions, and to recover the database. Oracle Database Configuration Assistant creates rollback segment identified by rbsthread_number where thread is the thread ID of the node, and number is the number (1 or 2) of the rollback segment. For example, the first node in a cluster would have rollback segments of rbs1_1 and rbs1_2. Rollback segments are indicated by the ROLLBACK_SEGMENTS parameter in the initsid.ora file.
Thread ID	Specifies the redo thread that is to be used by the instance. Any available redo thread number can be used, but an instance cannot use the same thread number as another instance. Also, an instance cannot start when its redo thread is disabled. An instance cannot mount a database if the thread is used by another instance or if the thread is disabled. The thread starts at 1 node for the first instance in the cluster, and is incremented by 1 for the next instance, and so on. Redo thread numbers are used in the naming of redo log files, which record changes made to data. When redo log files are generated, they include the thread, allowing you to easily identify a particular node's redo log files. Threads are indicated by the THREAD parameter in the initsid.ora file.

If the database name is op and the thread IDs are 1, 2 and 3, then the SID for each node is:

Thread ID	SID
1	op1
2	op2
3	op3

If the DB_NAME is op and the thread IDs are 1, 2 and 3, then the instance elements are affected in the following manner:

Computer	Thread ID	SID	INSTANCE_NAME	ROLLBACK_SEGMENT name
Computer A	1	op1	op1	rbs1-1, rbs1-2
Computer B	2	op2	op2	rbs2-1, rbs2-2
Computer C	3	op3	op3	rbs3_1, rbs3_2

Using Manual Methods

Oracle Database Configuration is intended primarily for new installations and configurations. If you already have existing scripts that are customized to your environment, it is advisable to create the database manually after installation, as described in "Use Manual Methods".