High Availability : Design, Techniques and Processes

by Piedad, Floyd; Hawkins, Michael W.

• ISBN13: 9780130962881
• ISBN10: 0130962880
• Edition: 1st
• Format: Hardcover
• Copyright: 2000-12-18
• Publisher: Prentice Hall

Summary

Introduction

Back when the mainframe ruled the world, information technology (IT) practitioners quickly learned the value of a well-managed system. They understood the value of managing problems, changes, and other issues confronting large, mission-critical computer systems running an organization's most sensitive business functions.

When the popularity of mainframes waned in favor of less costly midrange and PC systems, IT organizations were caught in the frenzy of developing and deploying new business applications with breakneck speed. Suddenly, more computing power was available to end users, who wanted to accomplish more with it than ever before. The corporate information system grew in scope, use, and importance, with no end in sight.

Now that the dust has settled somewhat, both the IT organization and the leaders of the business recognize that an unmanaged state-of-the-art computer system can be as bad as having none at all. Symptoms of this problem with unmanaged systems manifest themselves in ballooning IT costs, overworked and demoralized IT staff, and user dissatisfaction.

This book demonstrates how to deliver maximum system availability and manageability throughout a computer system's lifecycle, from design through implementation and maintenance. We review every key technique for simplifying the management and maintenance of computer systems - including redundancy, standardization, backups, and many more. We discuss practical means of implementing these techniques to make your current and future systems far less prone to outages.

We cover technical and management issues, since you cannot achieve long-term system availability and manageability solutions without addressing both. We have written this book to benefit everyone in the IT organization. Technical staff will find practical operational solutions that can be implemented immediately. IT management will gain a better perspective of the end-to-end and interrelated requirements of running an IT shop. And Chief Information Officers (CIOs) and other senior IT executives will find forward-looking strategies for enhancing the IT infrastructure and its contribution to the corporate bottom line.

You can manage systems better if you design them with high systems availability in mind. This book will show you how to address your system availability problems, from start to finish.

Author Biography

Floyd Piedad is President of AKAsia Services Corporation based in Manila, the Philippines. Michael Hawkins is an IT management consultant

Table of Contents

Introduction	xv
Acknowledgments	xvi
Today's Computing Environment	1	(12)
Complexity, Complexity, Complexity	1	(4)
Multiple Technologies and Protocols	1	(1)
Multiple Vendors	2	(1)
Varied Users	3	(1)
Multiple Locations	3	(1)
Rapid Change	4	(1)
Greater Business Demands	4	(1)
A Daunting Environment To Work In	4	(1)
The Total Cost of Ownership Issue	5	(7)
Total Cost of Ownership Defined	5	(2)
Industry TCO Estimates	7	(1)
What TCO Studies Reveal	7	(2)
The Underlying Reason for High TCO	9	(1)
A Typical Scenario: Choosing Office Systems	10	(1)
Availability as the Most Significant Contributor to TCO	11	(1)
Summary	12	(1)
Achieving Higher Availability	13	(12)
Determining User Availability Requirements	13	(3)
The Service Level Agreement	13	(1)
Helping Users Identify Their Availability Requirements	14	(2)
Availability Levels and Measurements	16	(7)
High Availability Level	16	(1)
Continuous Operations Level	17	(1)
Continuous Availability	17	(1)
Quantifying Availability Targets	17	(2)
Availability: A User Metric	19	(3)
Measuring End-To-End Availability	22	(1)
Summary	23	(2)
Planning for System Availability	25	(6)
Identifying System Components	25	(3)
Addressing Critical Components	28	(1)
The Four Elements of Availability	28	(1)
Summary	29	(2)
Preparing for Systems Management	31	(10)
Processes, Data, Tools, and Organization	31	(2)
Systems Management in the PC World (or the Lack of It)	33	(1)
IT Organizations: Away from Centralization, Then Back Again	33	(1)
Understanding the Systems To Manage	34	(2)
The Basics of Management: Five Phases	36	(2)
Setting Objectives	36	(1)
Planning	37	(1)
Execution	37	(1)
Measurement	37	(1)
Control	38	(1)
Identifying the Systems Management Disciplines	38	(3)
Implementing Service-Level Management	41	(52)
Service-Level Management	41	(8)
Process Requirements	41	(4)
Data and Measurement Requirements	45	(1)
Organization Requirements	46	(1)
Tools Requirements	47	(1)
Benefits of Service-Level Management	48	(1)
Problem Management	49	(9)
Process Requirements	49	(4)
Data Requirements	53	(2)
Organization Requirements	55	(1)
Tools Requirements	56	(2)
Benefits of Problem Management	58	(1)
Change Management	58	(7)
Process Requirements	59	(3)
Data Requirements	62	(1)
Organization Requirements	63	(1)
Tools Requirements	64	(1)
Benefits	64	(1)
Security Management	65	(12)
Process Requirements	65	(5)
Data Requirements	70	(3)
Organization Requirements	73	(2)
Tools Requirements	75	(1)
Benefits	76	(1)
Asset and Configuration Management	77	(8)
Process Requirements	77	(4)
Data Requirements	81	(2)
Organization Requirements	83	(1)
Tools Requirements	84	(1)
Availability Management	85	(8)
Process Requirements	86	(2)
Data Requirements	88	(1)
Organization Requirements	89	(1)
Tools Requirements	90	(1)
Benefits	91	(2)
From Centralized to Distributed Computing Environments	93	(12)
Systems Management Disciplines	93	(1)
The Centralized Computing Environment	94	(1)
The Distributed Computing Environment	95	(1)
Systems Management in Today's Computing Environment	96	(4)
Defining Appropriate Functions and Control	96	(1)
Choosing a Deployment Strategy	97	(3)
Developing a Deployment Strategy	100	(3)
Management by Exception	100	(1)
Policy-Based Management	101	(1)
Standardization of Performance Data	102	(1)
Accountability of the Distributed Systems Manager	102	(1)
Central Definition of Systems Management Architectures	102	(1)
Process Ownership	103	(1)
Summary	103	(2)
Techniques That Address Multiple Availability Requirements	105	(56)
Redundancy	105	(5)
Hardware Redundancy Examples	106	(2)
Software Redundancy Examples	108	(1)
Environmental Redundancy Example	109	(1)
Critical Success Factors	110	(1)
Backup of Critical Resources	110	(7)
Methods of Backup	111	(1)
Hardware Backup Examples	112	(1)
Software Backup Examples	112	(2)
IT Operations Backup Examples	114	(1)
Critical Success Factors	115	(2)
Clustering	117	(5)
Comparing Clustering and Redundancy	117	(2)
Hardware and Software Clustering Examples	119	(2)
IT Operations Clustering Examples	121	(1)
Environmental Clustering Examples	121	(1)
Critical Success Factors	121	(1)
Fault Tolerance	122	(3)
Hardware Fault Tolerance Examples	123	(1)
Software Fault Tolerance Examples	124	(1)
Environmental Fault Tolerance Examples	125	(1)
Critical Success Factors	125	(1)
Isolation or Partitioning	125	(6)
Hardware Isolation Examples	126	(1)
Software Isolation Examples	127	(1)
Other Benefits of Isolation	128	(1)
Critical Success Factors	129	(2)
Automated Operations	131	(7)
Console and Network Operations Examples	133	(1)
Workload Management Examples	134	(1)
System Resource Monitoring Examples	134	(1)
Problem Management Applications	135	(1)
Distribution of Resources Example	135	(1)
Backup and Restore Examples	136	(1)
Critical Success Factors	136	(2)
Access Security Mechanisms	138	(12)
Steps to Secure Access	139	(2)
Types of Security	141	(5)
Password Management	146	(2)
Critical Success Factors	148	(2)
Standardization	150	(8)
Hardware Standardization Examples	151	(1)
Software Standardization Examples	152	(1)
Network Standardization Examples	153	(1)
Processes and Procedures Standardization Examples	153	(1)
Naming Standardization Examples	154	(1)
Critical Success Factors	155	(2)
Transitioning to Standardization	157	(1)
Summary	158	(3)
Special Techniques for System Reliability	161	(22)
The Use of Reliable Components	161	(10)
Techniques for Maximizing Hardware Component Reliability	161	(3)
Techniques for Maximizing Software Component Reliability	164	(4)
Personnel-Related Techniques for Maximizing Reliability	168	(1)
Environment-Related Techniques for Maximizing Reliability	169	(1)
Some Reliability Indicators for Suppliers	170	(1)
Programming to Minimize Failures	171	(6)
Correctness	171	(2)
Robustness	173	(1)
Extensibility	174	(2)
Reusability	176	(1)
Implement Environmental Independence Measures	177	(3)
Use Power Generators	178	(1)
Use Independent Air-Conditioning Units	178	(1)
Use Fire Protection Systems	178	(1)
Use Raised Flooring	179	(1)
Install Equipment Wheel Locks	179	(1)
Locate Computer Room on the Second Floor	179	(1)
Utilize Fault Avoidance Measures	180	(1)
Analyzing Problem Trends and Statistics	180	(1)
Use of Advanced Hardware Technologies	180	(1)
Use of Software Maintenance Tools	181	(1)
Summary	181	(2)
Special Techniques for System Recoverability	183	(6)
Automatic Fault Recognition	183	(2)
Parity Checking Memory	183	(1)
ECC Memory	184	(1)
Data Validation Routines	184	(1)
Fast Recovery Techniques	185	(1)
Minimizing Use of Volatile Storage Media	186	(1)
Regular Database Updates to Central Storage	186	(1)
Automatic File-Save Features	186	(1)
Summary	187	(2)
Special Techniques for System Serviceability	189	(8)
Online System Redefinition	189	(1)
Add or Remove I/O Devices	189	(1)
Selectively Power Down Subsystems	190	(1)
Commit or Reject Changes	190	(1)
Informative Error Messages	190	(3)
Use Standard Corporate Terminology	191	(1)
Adopt Terms Already Used by Common Applications	191	(1)
Tell What, Why, Impact, and How	191	(1)
Implement Context-Sensitive Help	192	(1)
Give Options for Viewing More Detailed Error Information	192	(1)
Make Error Information Available After the Error Has Been Cleared	193	(1)
Complete Documentation	193	(2)
Have a Manual of Operations on Hand	193	(1)
Write Basic Problem Isolation and Recovery Guides	194	(1)
Provide System Configuration Diagrams	194	(1)
Label Resources	195	(1)
Provide a Complete Technical Library	195	(1)
Installation of Latest Fixes and Patches	195	(1)
Summary	196	(1)
Special Techniques for System Manageability	197	(14)
Use Manageable Components	197	(6)
Simple Network Management Protocol (SNMP)	199	(1)
Common Management Information Protocol (CMIP)	200	(1)
Desktop Management Interface (DMI)	201	(1)
Common Information Management Format (CIM)	202	(1)
Wired for Management (WfM)	202	(1)
Management Applications	203	(6)
Systems Management Issues	204	(1)
Automated Systems Management Capabilities	205	(1)
System Management Applications and Frameworks	206	(3)
Educate IS Personnel on Systems Management Disciplines	209	(1)
Business Value of the Information System	209	(1)
Value of Systems Management Disciplines	209	(1)
Principles of Management	209	(1)
Basic Numerical Analysis Skills	210	(1)
Summary	210	(1)
All Together Now	211	(8)
The Value of Systems Management Disciplines	211	(1)
Which One First?	212	(1)
Analyze Outages	213	(1)
Identify Single Points of Failure	214	(1)
Exploit What You Have	215	(1)
An Implementation Strategy	215	(1)
Summary	216	(3)
Appendix A Availability Features of Selected Products	219	(36)
Availability Features of Selected Operating Systems	219	(22)
Availability Features of Novell NetWare	220	(4)
Availability Features of Sun Solaris 8	224	(4)
Availability Features of AIX	228	(3)
Availability Features of Microsoft Windows 2000 Server and Professional	231	(8)
Availability Features of IBM OS/400	239	(2)
Availability Features of Selected Hardware Components	241	(12)
Availability Features of IBM S/390 Integrated Server	241	(2)
Availability Features of the IBM AS/400 Midrange System	243	(5)
Availability Features of the IBM RS/6000	248	(2)
Availability Features of Compaq Proliant Servers	250	(3)
Availability Features of Selected Software Components	253	(2)
Availability Features of the Oracle 8i Database	253	(2)
Index	255

Excerpts

Introduction Back when the mainframe ruled the world, information technology (IT) practitioners quickly learned the value of a well-managed system. They understood the value of managing problems, changes, and other issues confronting large, mission-critical computer systems running an organization's most sensitive business functions. When the popularity of mainframes waned in favor of less costly midrange and PC systems, IT organizations were caught in the frenzy of developing and deploying new business applications with breakneck speed. Suddenly, more computing power was available to end users, who wanted to accomplish more with it than ever before. The corporate information system grew in scope, use, and importance, with no end in sight. Now that the dust has settled somewhat, both the IT organization and the leaders of the business recognize that an unmanaged state-of-the-art computer system can be as bad as having none at all. Symptoms of this problem with unmanaged systems manifest themselves in ballooning IT costs, overworked and demoralized IT staff, and user dissatisfaction. This book demonstrates how to deliver maximum system availability and manageability throughout a computer system's lifecycle, from design through implementation and maintenance. We review every key technique for simplifying the management and maintenance of computer systems - including redundancy, standardization, backups, and many more. We discuss practical means of implementing these techniques to make your current and future systems far less prone to outages. We cover technical and management issues, since you cannot achieve long-term system availability and manageability solutions without addressing both. We have written this book to benefit everyone in the IT organization. Technical staff will find practical operational solutions that can be implemented immediately. IT management will gain a better perspective of the end-to-end and interrelated requirements of running an IT shop. And Chief Information Officers (CIOs) and other senior IT executives will find forward-looking strategies for enhancing the IT infrastructure and its contribution to the corporate bottom line. You can manage systems better if you design them with high systems availability in mind. This book will show you how to address your system availability problems, from start to finish.