Document revision date: 30 March 2001

Chapter 4
4	Evaluating System Resources
4.1	Prerequisites
4.2	Guidelines
4.3	Collecting and Interpreting Image-Level Accounting Data
4.3.1	Guidelines
4.3.2	Enabling and Disabling Image-Level Accounting
4.3.3	Generating a Report
4.3.4	Collecting the Data
4.4	Creating, Maintaining, and Interpreting MONITOR Summaries
4.4.1	Types of Output
4.4.2	MONITOR Modes of Operation
4.4.3	Creating a Performance Information Database
4.4.4	Saving Your Summary Reports
4.4.5	Customizing Your Reports
4.4.6	Report Formats
4.4.7	Using MONITOR in Live Mode
4.4.8	More About Multifile Reports
4.4.9	Interpreting MONITOR Statistics
Chapter 5
5	Diagnosing Resource Limitations
5.1	Diagnostic Strategy
5.2	Investigating Resource Limitations
5.2.1	Memory Limitations
5.2.2	I/O Limitations
5.2.3	CPU Limitations
5.3	After the Preliminary Investigation
5.3.1	Observing the Tuned System
5.3.2	Obtaining a Listing of System Current Values
Chapter 6
6	Managing System Resources
6.1	Understanding System Responsiveness
6.1.1	Detecting Bottlenecks
6.1.2	Balancing Resource Capacities
6.2	Evaluating Responsiveness of System Resources
6.3	Improving Responsiveness of System Resources
Chapter 7
7	Evaluating the Memory Resource
7.1	Understanding the Memory Resource
7.1.1	Working Set Size
7.1.2	Locality of Reference
7.1.3	Obtaining Working Set Values
7.1.4	Displaying Working Set Values
7.2	Evaluating Memory Responsiveness
7.2.1	Page Faulting
7.2.1.1	Hard and Soft Page Faults
7.2.1.2	Secondary Page Cache
7.2.2	Swapping and Swapper Trimming
7.3	Analyzing the Excessive Paging Symptom
7.3.1	What Is Excessive Paging?
7.3.2	Guidelines
7.3.3	Excessive Image Activations
7.3.4	Characterizing Hard Versus Soft Faults
7.3.5	System Page Faulting
7.3.6	Page Cache Is Too Small
7.3.7	Saturated System Disk
7.3.8	Page Cache Is Too Large
7.3.9	Small Total Working Set Size
7.3.10	Inappropriate WSDEFAULT, WSQUOTA, and WSEXTENT Values
7.3.10.1	Learning About the Process
7.3.10.2	Obtaining Process Information
7.3.11	Ineffective Borrowing
7.3.12	AWSA Might Be Disabled
7.3.13	AWSA Is Ineffective
7.3.13.1	AWSA Is Not Responsive to Increased Demand
7.3.13.2	AWSA with Voluntary Decrementing Enabled Causes Oscillations
7.3.13.3	AWSA Shrinks Working Sets Too Quickly
7.3.13.4	AWSA Needs Voluntary Decrementing Enabled
7.3.13.5	Swapper Trimming Is Too Vigorous
7.4	Analyzing the Swapping Symptom
7.4.1	Detecting Harmful Swapping
7.4.2	Investigating Harmful Swapping
7.4.3	Causes of Harmful Swapping
7.4.4	Why Processes Consume Unreasonable Amounts of Memory
7.4.5	Large, Compute-Bound Processes
7.4.6	Large Waiting Processes not Outswapped
7.4.7	Too Many Competing Processes
7.4.8	Borrowing Is Too Generous
7.4.9	Swapper Trimming Is Ineffective
7.4.10	Excessively Large Working Sets
7.4.11	Disk Thrashing Occurs
7.4.12	System Swaps Rather Than Pages
7.4.13	Demand Exceeds Available Memory
7.5	Analyzing the Limited Free Memory Symptom
7.5.1	Reallocating Memory
7.6	MONITOR Statistics for the Memory Resource
Chapter 8
8	Evaluating the Disk I/O Resource
8.1	Understanding the Disk I/O Resource
8.1.1	Components of a Disk Transfer
8.1.2	Disk Capacity and Demand
8.1.2.1	Seek Capacity
8.1.2.2	Data Transfer Capacity
8.1.2.3	Demand
8.2	Evaluating Disk I/O Responsiveness
8.2.1	Disk I/O Operation Rate
8.2.2	I/O Request Queue Length
8.2.3	Disk I/O Statistics for MSCP Served Disks
8.3	Disk or Tape Operation Problems (Direct I/O)
8.3.1	Software and Hardware Solutions
8.3.2	Determining I/O Rates
8.3.3	Abnormally High Direct I/O Rate
8.3.4	Paging or Swapping Disk Activity
8.3.5	Reduce I/O Demand or Add Capacity
8.4	Terminal Operation Problems (Buffered I/O)
8.4.1	Detecting Terminal I/O Problems
8.4.2	High Buffered I/O Count
8.4.3	Operations Count
8.4.4	Excessive Kernel Mode Time
8.5	MONITOR Statistics for the I/O Resource

	privacy and legal statement
6491PRO_CONTENTS_001.HTML