Oracle Real Application Clusters

The chapters are written to follow one another in a logical fashion by introducing topics in the earlier chapters and building on the technology. Thus it is advised that you read the chapters in order. Even if you have worked with clustered databases, you will certainly find a nugget or two that may be new to you. For the experienced reader, the book also highlights wherever applicable the new features introduced in Oracle 9i, Oracle 9i Release 2, and Oracle 10g. The book's four parts are outlined below.

Part I: Concepts

Information in this part is to get the reader's feet wet. This part contains three chapters.

• Chapter 1: Provides an overview of the modern business requirements and the criticality of these requirements in today's Internet-based business.

• Chapter 2: Focuses on the hardware concepts, the various types of hardware technologies such as SMP, MPP, NUMA, and the clustered SMP. The chapter also discusses the various components, such as the cluster interconnect technology and the storage systems, including various types of RAID technology.

• Chapter 3: Discusses the basic Oracle concepts, starting with what an instance is, then covering the various background and foreground processes, including the various types of files and tablespaces that Oracle supports and uses. Data block dump analysis and data block structure, including transaction layers, are discussed.

Part II: Architecture

Once the reader's feet are wet with regard to the concepts, the second part steps into the clustered database technology starting with the RAC architecture in Chapter 4. This part contains six chapters.

• Chapter 4: Takes an in-depth look into the RAC architecture, the various additional background and foreground processes required by RAC, their functions, and how they work together in clustered database architecture. The role of the GCS, GES, and GRD is covered a great detail.

• Chapter 5: Discusses the transaction management principles in general, the transaction management in a single instance configuration, and the transaction management in a RAC environment. This chapter discusses, using scenarios, the architecture behind this configuration, how the sharing of data occurs between nodes, and the sharing of data when the cluster has more than two nodes. Through examples, this chapter will explain how cache fusion helps the lock activity between the shared rows, the various lock levels, and how locks are handled, based on requests received from processes on various nodes participating in the cluster. It details the discussions concerning the cache fusion behavior in a transaction; various scenarios of clustered transaction management include the various states of GCS operation such as the PI and XI states of a block.

• Chapter 6: Discusses the concepts of parallel processing. The advantages of using parallel processing and the appropriate usage of this feature in today's business computing. Concepts such as working of parallel processing, parallel processing configuration scenarios, problems encountered in using parallel processing, tuning statements for parallel processing, and advantages of using parallel processing are also addressed.

• Chapter 7: Discusses the various features of the Oracle product and the technology to be considered in designing databases for the RAC environment. The chapter discusses various types of indexes, data partitioning and materialized views, and the advantages and positive effect of the technology for a RAC environment.

• Chapter 8: Covers the various installation and configuration steps required for RAC implementation. The chapter covers both the DBCA utility for database creation and the manual method, including configuration of the GSD, SRVCONFIG, and SRVCTL utilities, and listener configuration.

• Chapter 9: Illustrates, lists, and describes the various RAC-related data dictionary views and parameters

Part III: Operations

After a complete understanding of the architecture, the next step is to understand its operations, taking a step deeper and getting to the bottom of things. This part contains eight chapters.

• Chapter 10: Describes the availability and scalability features of RAC, including transparent application failover, Oracle Real Application Cluster Guard, fast reconfiguration, and inclusion and exclusion of instances. It also discusses how to configure TAF using the tnsnames file and by making OCI-based calls directly from a Java application.

• Chapter 11: Describes the steps involved during a migration process. Both migrating from a single instance to a clustered database environment and migrating from OPS to RAC are covered.

• Chapter 12: Covers the backup features available, including implementation and configuration of RMAN. There are extensive discussions on the various recovery operations, illustrated with trace file analysis, which includes LMON, SMON, and alert log files.

• Chapter 13: This is the first of three chapters covering a very important subject, namely performance tuning of a RAC cluster. This chapter discusses the various tools and utilities available such as STATSPACK, SQL_TRACE, Oracle Enterprise Manager (OEM), and methods of analysis by understanding the internals of Oracle, such as using the Oracle Event and Wait interface, are discussed in detail.

• Chapter 14: Tuning starts with a single instance; this chapter discusses in detail tuning of a single instance of the database. Starting with a tuning methodology, the chapter approaches tuning from the top down, tuning the application, followed by the instance, and then the database. In this chapter the various wait interface views are drilled down and analyzed to solve specific performance issues. The chapter discusses capturing session level statistics for a thorough analysis of the problem.

• Chapter 15: After tuning a single instance of the database, the next step is to tune the cluster. This chapter discusses this key piece to the puzzle. Cluster tuning is tuning of many instances working in a cohesive unit. This means there is sharing of data across middleware. This chapter discusses tuning the cluster interconnect, the shared storage subsystems, and other global cache management areas, including cache transfer and interinstance lock management.

• Chapter 16: Discusses the various diagnostic options available for troubleshooting an Oracle database. This includes discussions on common error messages such as ORA-600 and ORA-7445, how to troubleshoot these errors first, and how to gather information to help Oracle support diagnose the problem quickly. This chapter also discusses the Flash Freeze feature, ORADEBUG, and RDA.

• Chapter 17: Discusses implementation of RAC on other operating systems, the differences, and additional processes. Primarily, this chapter focuses on Linux and Windows operating systems. It discusses the clustered file system introduced in Oracle 9i Release 2 for the Windows operating system and the watchdog process for the Linux operating system.

Part IV: Appendices

• Appendix 1: Utilities

• Appendix 2: Scripts

• Appendix 3: References

• Appendix 4: Work Plan

Graphics used

The book uses graphic icons as shown in the following table to quickly locate information or to alert a reader regarding any important information.