The course starts with an unknown database that requires tuning. The lessons will proceed through the steps a DBA will perform to acquire the information needed to identify problem areas, to diagnose common problems, and remedy those problems. The methodology used in the practices is primarily reactive. After configuring monitoring tools, and reviewing the available reports, the student will be presented with the Oracle architecture based on the SQL statement processing  of SELECT and DML.

The SQL tuning section assumes that the DBA has little or no ability to change the code. The DBA will influence the SQL performance with available tools. The DBA will be introduced to various methods of identifying the SQL statements that require tuning, and the diagnostic tools needed to find ways to change the performance. This will include the use of statistics, outlines, and profiles to influence the optimizer, adding and rebuilding indexes, and using the SQL Advisors. A major task of DBA’s is to maintain SQL performance across changes. This course introduces the DB Replay, and SQL Performance Analyzer tools to help the DBA test and minimize the impact of change.

Instance tuning uses the same general method of observing a problem, diagnosing the problem, and implementing a solution. The instance tuning lessons cover the details of major tunable components and describe how you can influence the instance behavior. For each lesson, we will examine the relevant components of the architecture. The class only discusses the architecture to the level required to understand the symptoms and solutions. More detailed explanations are left to other courses, reference material, and the Oracle documentation.

The last lesson of this course is a recap of the best practices discovered in the previous lessons, and miscellaneous recommendations. The goal is to finish the course with a best practices list for students to take away.

• Use the Oracle Database tuning methodology appropriate to the available tools
• Utilize database advisors to proactively tune an Oracle Database Instance
• Use the tools based on the Automatic Workload Repository to tune the database
• Diagnose and tune common SQL related performance problems
• Diagnose and tune common Instance related performance problems
• Use Enterprise Manager performance-related pages to monitor an Oracle Database

• Database Administrators
• Support Engineer
• Technical Consultant

• Oracle Database 11g: Administration Workshop II Release 2
• Oracle Database 11g: Administration Workshop I Release 2

Using Database Replay
The Big Picture & System Architecture
Capture & Replay Considerations
Replay Options & Analysis
Database Replay Workflow in Enterprise Manager
Packages and Procedures
Data Dictionary Views: Database Replay
Database Replay: PL/SQL Example
Calibrating Replay Clients

Tuning the Shared Pool
Shared Pool Architecture & Operation
The Library Cache & Latch and Mutex
Diagnostic Tools for Tuning the Shared Pool
Avoiding Hard & Soft Parses
Sizing the Shared Pool & Avoiding Fragmentation
Data Dictionary Cache & SQL Query Result Cache
UGA and Oracle Shared Server
Large Pool & Tuning the Large Pool

Tuning the Buffer Cache
Oracle Database Architecture: Buffer Cache
Database Buffers
Buffer Hash Table for Lookups
Working Sets
Buffer Cache Tuning Goals and Techniques
Buffer Cache Performance Symptoms & Solutions
Automatically Tuned Multiblock Reads
Flushing the Buffer Cache (for Testing Only)

Tuning PGA and Temporary Space
SQL Memory Usage & Performance Impact
SQL Memory Manager
Configuring Automatic PGA Memory & Setting PGA_AGGREGATE_TARGET Initially
Monitoring & Tuning SQL Memory Usage
PGA Target Advice Statistics & Histograms
Automatic PGA and Enterprise Manager & Automatic PGA and AWR Reports
Temporary Tablespace Management: Overview & Monitoring Temporary Tablespace
Temporary Tablespace Shrink & Tablespace Option for Creating Temporary Table

Automatic Memory Management
Oracle Database Architecture, Dynamic SGA & Memory Advisories
Granule & Manually Adding Granules to Components
Increasing the Size of an SGA Component, SGA Sizing Parameters & Manually Resizing Dynamic SGA Parameters
Automatic Shared Memory Management & Memory Broker Architecture
Behavior of Auto-Tuned & Manually TunedSGA Parameters
Using the V$PARAMETER View & Resizing SGA_TARGET
Disabling, Configuring & Monitoring Automatic Shared Memory Management (ASMM)
Automatic Memory Management