This book is intended for system administrators and operators who are responsible for maintaining the integrity and availability of mission-critical UNIX systems. The book provides a description of the fault monitoring tools and techniques available for UNIX servers, including systems that are configured as high availability clusters.
This book can therefore be a handy quick reference for an operator trying to troubleshoot a problem in the customer environment, by pointing out where to find key diagnostic messages and describing how to take recovery actions.
A system administrator responsible for the initial configuration and administration of UNIX systems will also find this book useful because it describes the procedures to follow to set up the appropriate levels of system monitoring. The product descriptions can also help in making purchasing decisions as the customer determines the appropriate amount of event monitoring needed in their environment.
An overview of the tasks performed by an operator is provided, with details on how events are received and processed. The remainder of the book focuses on the types of events that can be received, how they are detected, how operators receive event notifications, and how problems can be investigated and recovery performed. The goal is to introduce the necessary tools, but not to show how every possible problem can be solved.
This book provides numerous descriptions of how fault management tools and products can be used to solve a variety of problems. Many of the chapters are focused on specific computer components, such as disks or databases, to be helpful to operators with specific roles. Here is a description of the individual chapters:
Chapter 1, "Analyzing the Role of System Operators," describes the tasks performed by a system operator and the evolution of fault management.
Chapter 2, "Enumerating Possible Events," describes the various types of events that are interesting to monitor on a UNIX system.
Chapter 3, "Using Monitoring Frameworks," describes monitoring frameworks and the administrative tasks that must be done before they can be used.
Chapter 4, "Monitoring the System," describes the tools and products used to monitor the UNIX server.
Chapter 5, "Monitoring the Disks," describes the tools and products used to monitor external disk devices.
Chapter 6, "Monitoring the Network," provides an overview of the many tools available for detecting problems and events related to the use of the network.
Chapter 7, "Monitoring the Application," describes methods for monitoring the response times and availability of critical applications.
Chapter 8, "Monitoring the Database," focuses specifically on tools to detect problems and events related to database usage.
Chapter 9, "Enterprise Management," discusses the problems with trying to deal with fault management for the large-scale customer enterprise.
Chapter 10, "UNIX Futures," discusses the future plans of some of the major UNIX system vendors in the area of fault management.
Appendix A, "Standards," describes fault management standards that have emerged and how you can benefit from them.
The Glossary contains the important terms used in the book, and their definitions.
Although it is assumed that most customers concerned about fault management will implement high availability solutions, this book does not describe how to create highly available computing environments. Readers needing additional information on high availability may check Hewlett-PackardÕs external Web site on high availability (hp/go/ha) or read Clusters for High Availability by Peter Weygant.
In general, this book does not discuss the configuration and installation of the hardware and software components of your UNIX system. You should rely on your vendorsÕ product manuals for this.
Many of the examples used in this book were created on HP-UX servers. Other UNIX platforms behave similarly, and we note when tools are supported only on certain UNIX platforms.