The outline: Oracle Database 11g: Performance Tuning DBA Release 2

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.

Learn To:

• Describe Oracle tuning methodology
• Use Oracle supplied tools for monitoring, and diagnosing SQL and Instance tuning issues
• Use database advisors to correct performance problems proactively
• Identify problem SQL statements & tune SQL performance problems
• Monitor the Instance Performance using Enterprise Manager
• Tune instance components, primarily using Instance parameters

Introduction

• This lesson introduces the Performance Tuning course objectives and agenda

Basic Tuning Tools

• Monitoring tools overview
• Enterprise Manager
• V$ Views, Statistics and Metrics
• Wait Events

Using Automatic Workload Repository

• Managing the Automatic Workload RepositoryCreate AWR Snapshots
• Real Time SQL Monitoring (a 11.1 feature new lesson in NF L-15)

Defining Problems

• Defining the Problem
• Limit the Scope & Setting the Priority
• Top SQL Reports
• Common Tuning Problems & Tuning During the Life Cycle
• ADDM Tuning Session
• Performance Versus Business Requirements
• Performance Tuning Resources & Filing a Performance Service Request
• Monitoring and Tuning Tools: Overview

Using Metrics and Alerts

• Metrics, Alerts, and Baselines
• Limitation of Base Statistics & Typical Delta Tools
• Oracle Database 11g Solution: Metrics
• Benefits of Metrics
• Viewing Metric History Information & Vsing EM to View Metric Details
• Statistic Histograms & Histogram Views
• Database Control Usage Model & Setting Thresholds
• Server-Generated Alerts, Creating and Testing an Alert & Metric and Alert Views

Using Baselines

• Comparative Performance Analysis with AWR Baselines
• Automatic Workload Repository Baselines
• Moving Window Baseline
• Baselines in Performance Page Settings & Baseline Templates
• AWR Baselines & Creating AWR Baselines
• Managing Baselines with PL/SQL & Baseline Views
• Performance Monitoring and Baselines & Defining Alert Thresholds Using a Static Baseline
• Using EM to Quickly Configure & Changing Adaptive Threshold Settings

Using AWR Based Tools

• Automatic Maintenance Tasks
• ADDM Performance Monitoring
• Active Session History: Overview

Monitoring an Application

• What Is a Service? Service Attributes & Service Types
• Creating Services & Managing Services in a Single-Instance Environment
• Everything Switches to Services.
• Using Services with Client Applications & Using Services with the Resource Manager
• Services and Resource Manager with EM & Using Services with the Scheduler
• Using Services with Parallel Operations & Metric Thresholds
• Service Aggregation and Tracing & Service Aggregation Configuration.
• Client Identifier Aggregation and Tracing & Service Performance Views

Identifying Problem SQL Statements

• SQL Statement Processing Phases & Role of the Oracle Optimizer
• Identifying Bad SQL, Real Time SQL Monitoring (a 11.1 feature new lesson in NF L-15) & TOP SQL Reports
• What Is an Execution Plan? Methods for Viewing Execution Plans & Uses of Execution Plans
• DBMS_XPLAN Package: Overview & EXPLAIN PLAN Command
• Reading an Execution Plan, Using the V$SQL_PLAN View & Querying the AWR
• SQL*Plus AUTOTRACE & SQL Trace Facility
• How to Use the SQL Trace Facility
• Generate an Optimizer Trace

Influencing the Optimizer

• Functions of the Query Optimizer, Selectivity, Cardinality and Cost & Changing Optimizer Behavior
• Using Hints, Optimizer Statistics & Extended Statistics
• Controlling the Behavior of the Optimizer with Parameters
• Enabling Query Optimizer Features & Influencing the Optimizer Approach
• Optimizing SQL Statements, Access Paths & Choosing an Access Path
• Join & Sort Operations
• How the Query Optimizer Chooses Execution Plans for Joins
• Reducing the Cost

Using SQL Performance Analyzer

• Real Application Testing: Overview & Use Cases
• SQL Performance Analyzer: Process & Capturing the SQL Workload
• Creating a SQL Performance Analyzer Task & SPA (NF Lesson 9) DBMS_SQLTUNE.CREATE_TUNING_TASK
• Optimizer Upgrade Simulation & SQL Performance Analyzer Task Page
• Comparison Report & Comparison Report SQL Detail
• Tuning Regressing Statements & Preventing Regressions
• Parameter Change Analysis & Guided Workflow Analysis
• SQL Performance Analyzer: PL/SQL Example & Data Dictionary Views

SQL Performance Management

• Maintaining SQL Performance and Optimizer Statistics & Automated Maintenance Tasks
• Statistic Gathering Options & Setting Statistic Preferences
• Restore Statistics
• Deferred Statistics Publishing: Overview & Example
• Automatic SQL Tuning: Overview
• SQL Tuning Advisor: Overview
• Using the SQL Access Advisor
• SQL Plan Management: Overview

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

Tuning Segment Space Usage

• Space and Extent Management & Locally Managed Extents
• How Table Data Is Stored & Anatomy of a Database Block
• Minimize Block Visits
• The DB_BLOCK_SIZE Parameter
• Small & Large Block Size: Considerations
• Block Allocation, Free Lists & Block Space Management with Free Lists
• Automatic Segment Space Management
• Migration and Chaining, Shrinking Segments & Table Compression: Overview

Tuning I/O

• I/O Architecture, File System Characteristics, I/O Modes & Direct I/O
• Bandwidth Versus Size & Important I/O Metrics for Oracle Databases
• I/O Calibration and Enterprise Manager, I/O Calibration and the PL/SQL Interface & I/O Statistics and Enterprise Manager
• Stripe and Mirror Everything
• Using RAID
• I/O Diagnostics
• Database I/O Tuning
• What Is Automatic Storage Management?

Performance Tuning Summary

• Best practices identified throughout the course
• Summarize the performance tuning methodology

Appendix B: Using Statspack

• Installing Statspack
• Capturing Statspack Snapshots
• Reporting with Statspack
• Statspack Considerations
• Statspack and AWR Reports
• Reading a Statspack Report
• Statspack and AWR