Tài liệu Module 1: Introduction to Windows Clustering docx

Contents
Overview 1
Defining Clustering Features 2
Introducing Application Architecture 4
Identifying Availability and Scalability
Requirements 6
Introducing Microsoft Windows 2000
Clustering 13
Comparing Network Load Balancing to
Cluster Service 18
Identifying the Application and Service
Environments 19
Discussion: Evaluating Business Scenarios 21
Review 27

Module 1: Introduction
to Windows Clustering

Information in this document is subject to change without notice. The names of companies,
products, people, characters, and/or data mentioned herein are fictitious and are in no way intended
to represent any real individual, company, product, or event, unless otherwise noted. Complying
with all applicable copyright laws is the responsibility of the user. No part of this document may
be reproduced or transmitted in any form or by any means, electronic or mechanical, for any
purpose, without the express written permission of Microsoft Corporation. If, however, your only
means of access is electronic, permission to print one copy is hereby granted.

Microsoft may have patents, patent applications, trademarks, copyrights, or other intellectual
property rights covering subject matter in this document. Except as expressly provided in any

written license agreement from Microsoft, the furnishing of this document does not give you any
license to these patents, trademarks, copyrights, or other intellectual property.

 2000 Microsoft Corporation. All rights reserved.

Microsoft, Active Directory, BackOffice, Jscript, PowerPoint, Visual Basic, Visual Studio, Win32,
Windows, Windows NT are either registered trademarks or trademarks of Microsoft Corporation
in the U.S.A. and/or other countries.

Other product and company names mentioned herein may be the trademarks of their respective
owners.

Program Manager: Don Thompson
Product Manager: Greg Bulette
Instructional Designers: April Andrien, Priscilla Johnston, Diana Jahrling
Subject Matter Experts: Jack Creasey, Jeff Johnson
Technical Contributor: James Cochran
Classroom Automation: Lorrin Smith-Bates
Graphic Designer: Andrea Heuston (Artitudes Layout & Design)
Editing Manager: Lynette Skinner
Editor: Elizabeth Reese
Copy Editor: Bill Jones (S&T Consulting)
Production Manager: Miracle Davis
Build Manager: Julie Challenger
Print Production: Irene Barnett (S&T Consulting)
CD Production: Eric Wagoner
Test Manager: Eric R. Myers
Test Lead: Robertson Lee (Volt Technical)
Creative Director: David Mahlmann
Media Consultation: Scott Serna

Illustration: Andrea Heuston (Artitudes Layout & Design)
Localization Manager: Rick Terek
Operations Coordinator: John Williams
Manufacturing Support: Laura King; Kathy Hershey
Lead Product Manager, Release Management: Bo Galford
Lead Technology Manager: Sid Benavente
Lead Product Manager, Content Development: Ken Rosen
Group Manager, Courseware Infrastructure: David Bramble
Group Product Manager, Content Development: Julie Truax
Director, Training & Certification Courseware Development: Dean Murray
General Manager: Robert Stewart


Module 1: Introduction to Windows Clustering iii

Instructor Notes
This module provides students with an overview of clustering technologies. The
module begins by defining the basic clustering features, introducing the
application architecture and identifying clustering solutions. The students are
then introduced to the Microsoft
® Windows® 2000 Advanced Server clustering
technologies.
After completing this module, students will be able to:
 Define clustering features.
 Define application architecture.
 Identify clustering technologies that can improve availability and scalability
in an enterprise system.
 Identify the available Microsoft clustering technologies.
 Identify the similarities and appropriate use of the clustering technologies.
 Identify the applications and services that can benefit from clustering

technologies.

Materials and Preparation
This section provides the materials and preparation tasks that you need to teach
this module.
Required Materials
To teach this module, you need the Microsoft PowerPoint® file 2087A.01.ppt
Preparation Tasks
To prepare for this module, you should:
 Read all of the materials for this module.
 Complete the labs.
 Study the review questions and prepare alternative answers to discuss.
 Anticipate questions that students may ask. Write out the questions and
provide the answers.

Presentation:
90 Minutes
iv Module 1: Introduction to Windows Clustering

Module Strategy
Use the following strategy to present this module:
 Defining Clustering Features
• Review and discuss each of the clustering feature definitions and clarify
any points of confusion for the students.
• Be prepared to discuss the Mean Time Between Failure concepts that are
presented in the topic Comparing Reliability and Availability. A white
board representation of hardware dependencies would be appropriate.
 Introducing Application Architecture
• This topic is an overview to the application architecture model.
Find out how knowledgeable the students are in each of the technologies

and identify whether students require more background information on a
given technology.
• Emphasize that the Application Architecture model is generic in nature.
Organizations typically customize the model to suit their unique
situations. Avoid spending time discussing implementation and
management philosophies.
• Be prepared to expand each level of the slide graphic and discuss the
application architecture of each.
 Identifying Availability and Scalability Solutions
• This topic describes how clustering can provide solutions for scalability
and availability issues.
• Discuss the importance of completing a risk assessment to determine if
clustering is a viable solution.
• Emphasize the importance of providing and maintaining a high levels of
availability and scalability for applications and data.
• Use the example provided to clearly demonstrate the concept of
measuring high availability.
 Introducing Microsoft Windows 2000 Clustering
• This topic provides a brief overview of each of the Windows 2000
Advanced Server clustering technologies.
• Emphasize the feature sets for each of the clustering technologies.
• Demonstrate where each technology can fit into the network
architecture.
 Comparing Network Load Balancing to Cluster Service
• This topic describes the benefits of both Cluster Service and Network
Load Balancing.
• Briefly review the benefits for each technology and how that applies to
several different business scenarios.
Module 1: Introduction to Windows Clustering v

 Identifying the Application and Service Environments
• This topic is an overview of the applications and services that are
compatible with a clustering solution.
• Emphasize the importance of identifying which applications and services
are appropriate for a clustering solution.
 Discussion: Evaluating Business Scenarios
• Review Scenario one and Scenario two with the students and have them
complete both scenarios.
• Although Scenario one is the simpler of the two scenarios, explain to the
students that they are expected to recognize where problems exist in the
scenario.
• Scenario two is much more complex. Explain to the students that they
are expected to consider:
• Use of Network Load Balancing, Component Load Balancing and
Cluster Service in both scenarios
• How applications and development processes will influence the
possible solutions
• How fault tolerance would be implemented within the applications
used
• Students may not have experience at this level. Be expected to lead
the discussion.


Module 1: Introduction to Windows Clustering 1

Overview
 Defining Clustering Features
 Introducing Application Architecture
 Identifying Availability and Scalability Requirements
 Introducing Microsoft Windows 2000 Clustering

 Comparing Network Load Balancing to Cluster Service
 Identifying the Application and Service Environments

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
As your organization’s business needs grow, you must be able to expand your
organization’s system capacity economically, avoid single points of failure and
quickly restore failed services and applications for users. Microsoft
®
Windows® 2000 Clustering enables you to provide availability, scalability, and
load balancing for applications and services.
This module describes the central concepts of Cluster service and Network
Load Balancing service, by providing a brief background of clustering
technologies and explaining what Windows 2000 Clustering provides.
In this course, a cluster is defined as a group of independent computers working
together as a single system. Microsoft clustering technologies provide the
functionality that is required to enable you to configure multiple computers as a
single logical system. In this module, you will learn the key benefits of
Microsoft Windows 2000 Clustering and how they apply within single and
multiple tier application architectures.
After completing this module, you will be able to:
 Define clustering features.
 Define application architectures.
 Identify clustering technologies that can improve availability and scalability
in an enterprise system.
 Identify the available Microsoft clustering technologies.
 Identify the similarities and appropriate use of the clustering technologies.
 Identify the applications and services that can benefit from clustering
technologies.

Topic Objective
To provide an overview of
the module topics and
objectives.
Lead-in
In this module, you will learn
the central concepts of
Windows 2000 Clustering
technologies.
2 Module 1: Introduction to Windows Clustering

Defining Clustering Features
 High Availability and Fault Tolerance
 Manageability
 Scalability
 Comparing Reliability and Availability

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
A working knowledge of a clustering solution begins with the definitions of
clustering features.
High Availability and Fault Tolerance
A system that is available whenever users want to use it and provides service
that meets a defined organizational standard is considered to have high
availability.
When a system or component in a cluster fails, the cluster software responds by
reallocating the resources from the failed system to the remaining systems in
the cluster, thereby ensuring that the system is providing high availability to
client/server applications and services.
Throughout this process, client communications with applications or services

usually continue with minimal interruption in service and Clustering provides a
single, virtual image of the server to clients. Most client software applications
will automatically recover from the broken connections with little or no
interruption to the user.
A fault tolerant solution is one that addresses performance by offering error-
free, nonstop availability, usually by keeping a backup of the primary system.
This backup system remains idle and unused until a failure occurs, which
makes this an expensive solution.
Manageability
Although manageability is not a key feature of clustering technologies, it allows
system administrators to perform all of the necessary functions of maintaining
the system by providing a single point of control. Administrators can access a
single point of control remotely or run tools that provide a view of the system
members, which allows control of the servers as a single logical entity.
Topic Objective
To define clustering
concepts.
Lead-in
A working knowledge of
Cluster service begins by
learning the definitions of
clustering features.
Delivery Tip
Emphasize the difference
between fault tolerance and
high availability. High
availability does not ensure
that a system is fault
tolerant.

Carefully review the Mean
Time Between Failure, use
the example provided to
demonstrate the concept.
Be prepared for questions
from the students regarding
geographical high
availability. For example,
there are two data centers,
one in Houston, Texas and
the other in San Ramon,
California. The students
might want to know how
availability is maintained
between the two data
centers if a catastrophic
event occurred.
Module 1: Introduction to Windows Clustering 3

Scalability
A system can be scaled up, scaled out, or scaled down.
 Scaling up. Achieved by adding more resources, such as memory,
processors, and disk drives to a system.
 Scaling out. Achieved by adding additional computers to deliver high
performance when the throughput requirements of an application exceed the
capabilities of an individual system.
 Scaling down. Achieved by reducing resources.

When the overall load exceeds the capabilities of the systems in a cluster, you
may need to add additional systems. You will find that clusters are highly

scalable; you can add CPU, input/output (I/O) storage, and application
resources incrementally to efficiently expand or contract capacity by
implementing one of the three types of scaling architectures.
Comparing Reliability and Availability
High availability and high reliability are at times used interchangeably, but
when considering complex systems, each can have a different meaning.
When designing products, for example a computer motherboard, there is a
failure rate defined for each component. The reliability number may be
expressed as Mean Time Between Failure (MTBF), which shows the measured
failure rate based on testing of the individual components.
The testing regime is usually a large number of components being tested in a
benign environment within their operating parameters, the aggregate run hours
without failure are used to ascertain the MTBF. Given the reliability figures of
all of the components, it is possible to calculate the probability of failure of the
motherboard within a given time. This MTBF number is a measure of the
reliability of the component and recognizes that all components will fail in
time. For example, disk drives may have an MTBF of 1x10
6
power hours.
A system with high availability is one where you expect that whenever you
want to use it, it is available to provide service meeting your defined standard.
So a computer system might be expected to be available 24 hours a day, 7 days
a week, 52 weeks a year; in other words, it can never stop working. There is a
distinct advantage to using high reliability components to build high availability
systems, because the probability of a failure is lower. However, you can build
high availability systems by using unreliable components, provided that you use
some fault-tolerant mechanism to maintain operation.
4 Module 1: Introduction to Windows Clustering




 Introducing Application Architecture
User Services
User Services
User Services
Business Services
Business Services
Business Services
Data Services
Data Services
Data Services
Two-Tier
Thin Client
Two-Tier
Thin Client
Two-Tier
Fat Client
Two-Tier
Fat Client
Three-Tier
Three-Tier
N-Tier
N-Tier
User Interface
Microsoft
Win32®
User Interface
Microsoft
Win32®
User Interface

Win32
Most Business
Logic
User Interface
Win32
Most Business
Logic
User Interface
Win32
Browser
User Interface
Win32
Browser
User Interface
Win32
Browser
DHTML, XML
User Interface
Win32
Browser
DHTML, XML
Business Logic
COM Objects
Business Logic
COM Objects
User Interface
ASP
Business Logic
COM Objects
User Interface

ASP
Business Logic
COM Objects
Storage
RDBMS
All Business
Logic (SP)
Storage
RDBMS
All Business
Logic (SP)
Storage
RDBMS
Min Business
Logic (SP)
Storage
RDBMS
Min Business
Logic (SP)
Storage
RDBMS
Min Business
Logic (SP)
Storage
RDBMS
Min Business
Logic (SP)
Storage
RDBMS
Min Business

Logic (SP)
Storage
RDBMS
Min Business
Logic (SP)

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
The application architecture defines how pieces of the application interact with
each other, and what functionality each piece is responsible for performing.
There are three main classes of application architecture that can be
characterized by the number of layers between the user and the data.
The three types of application architecture are two-tier, three-tier and n-tier,
where n can be three or more. The table demonstrates the user, business, and
data services layers in each of the application architectures. One of the benefits
of a three-tier or n-tier model is that applications are divided cleanly into
presentation, business logic, and data layers. This division results in enhanced
scalability and manageability, which can be improved by Windows Clustering
technologies.
Two-Tier
In a thin client, two-tiered model, the business logic is server-based and
typically consists of stored procedures in the database server. You must install
client code on every client accessing the application; the client code is
responsible for the user interface only.
In a fat client, two-tiered model, you must install client code on every client
accessing the application; the client code is responsible for the user interface
and most of the business logic. The database can still have stored procedures,
but the requirements for these procedures are reduced. This model requires that
more resources are available on the client.
Three-Tier

In a three-tiered model, the business layer or application layer lies between data
and client. This layer is responsible for both the application's business logic and
the overall management of business transactions. Often the application layer
will utilize object technologies.
Topic Objective
To introduce the
fundamentals of application
architecture.
Lead-in
Application architecture
defines how pieces of the
application interact with
each other, and what
functionality each piece is
responsible for performing.
Delivery Tip
Be prepared to expand all
three tiers of the model and
explain the applications and
services in each of them.
Key Point
One of the benefits of a
three-tier or n-tier model is
that applications are divided
cleanly into presentation,
business logic, and data
layers; the result will be
enhanced scalability and
manageability.
Module 1: Introduction to Windows Clustering 5

N-Tier
In an n-tier model, the user-services tier or first tier handles presentation of
information and interaction with the users. Some sources refer to this first tier
as presentation services, because some of the services that are performed in the
middle or business services tier of an application, such as authenticating users,
are also user services.
The business-services tier provides most of an application's functionality. This
tier handles the bulk of application-specific processing and enforces an
application's business rules. Business logic built into custom components
bridges the client environments and the data-services tier.
The data-services tier in an n-tier application can consist of data residing in
several different kinds of stores.
Although this split is conceptual, it can be mirrored in a real-world scenario by
implementing the data tier on a number of computers running a high
performance database, such as Microsoft SQL Server
™
; implementing the
business tier on a set of separate computers; and implementing the presentation
tier on yet another set of computers. When you complete the n-tier
implementation, you achieve redundancy and if a single computer fails, the
applications and services are available on the other computers. This
environment also addresses the need for scalability by allowing users to
incorporate different hardware.
Application Architecture Development Strategies
Developing applications for a Microsoft platform will typically use Microsoft
development tools and strategies. Current applications use the Windows
Distributed interNet Application Architecture (Windows DNA) strategies for
development and future development extend this using Microsoft .NET
Enterprise Servers.

Windows DNA
The Windows DNA model distributes an application in several layers, called
tiers, which often reside physically on different machines, emphasizing logical
distribution.
Microsoft developed Windows DNA as a way to fully integrate the Web with
the n-tier model of development. Windows DNA defines a framework for
delivering solutions that meet the demanding requirements of corporate
computing, the Internet, intranets, and global electronic commerce, while
reducing overall development and deployment costs.
Windows DNA architecture employs standard Windows-based services to
address the requirements of each tier in the multitiered solution: user interface
and navigation, business logic, and data storage.
Microsoft .NET
The core services of .NET are fulfilled by a set of strategies for the
development Internet-based applications. These core services include services
and development strategies for user identification, data storage, calendar
management, messaging, database, and many other services.
6 Module 1: Introduction to Windows Clustering



 Identifying Availability and Scalability Requirements
 Assessing Risks
 Scalability
 High Availability

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
As a system administrator planning to expand your system’s capacity, you may
be required to make commitments to expensive high-end servers that provide

space for additional CPUs, drives, and memory. By using a clustering
technology solution, you will be able to incrementally add smaller, standard
systems as needed to meet overall processing power requirements.
Clustering solutions are ideal when you need more system processing power or
high availability. For example, you would consider using a clustering solution
for an Internet server-based program supporting mission-critical applications,
such as financial transactions, database access, corporate intranets, and other
key functions that must run 24 hours a day, 7 days a week.
Implementing a clustering solution makes it possible for you to share a
computing load over several computer systems, without the users needing to
know that more than one computer is involved. If any component in the system
(hardware or software) fails, the user will not lose access to the service or
application.
Topic Objective
To identify how clustering is
a solution for availability and
scalability issues.
Lead-in
As a system administrator
planning to expand your
system’s capacity, you may
be required to make
commitments to expensive
high-end servers that
provide space for additional
CPUs, drives, and memory.
Delivery Tip
This topic is meant to
provide just an overview of
scalability and high

availability. Advise the
students that each of these
features will be covered in
much more detail in the
Server Cluster and Network
Load Balancing cluster
modules.
Module 1: Introduction to Windows Clustering 7

Assessing Risks
Client
Client
Client
Client
Client
Client
Client
Client
Router
Router
Server
Server
Power
Power
Performing a Risk Audit
Performing a Risk Audit

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
A risk audit helps you to identify system risk; it also helps to determine if

clustering is an appropriate solution to reduce the risk. More specifically it
helps to identify where you can use clustering to eliminate single points of
failure and maintain availability.
Identifying Risks
When you identify risks, you identify the possible failures that can interrupt
access to resources. A single point of failure is any component in your
environment that would block data or applications if it failed. A single point of
failure can be caused by hardware, software, or external dependencies, such as
power supplied by a utility company and dedicated wide area network (WAN)
lines.
In general, you provide improved reliability when you minimize the number of
single points of failure in your environment. Maximum reliability is provided
by mechanisms that maintain service when a failure occurs by providing fault
tolerance.
Topic Objective
To identify possible points of
failure before implementing
a clustering solution.
Lead-in
A risk audit identifies
network risk and helps to
determine if clustering is an
appropriate solution to
reduce risk.
Delivery Tip
Emphasize the importance
of identifying single points of
failure while doing a risk
audit prior to installing the
Network Load Balancing

driver.
8 Module 1: Introduction to Windows Clustering

Performing a Risk Audit
The following table lists some of the more commonly encountered points of
failure.
Point of failure Cluster service solution Possible other solutions

Network component, such
as a hub or router
None Spare components or
redundant routes
Power failure None Uninterruptible power
supply (UPS)
Server hardware, such as
CPU, memory, or network
card
Failover process of taking
resources offline on one
node and bringing them
back online on another
node
None
Disk – non shared Failover None
Disk – shared None Redundant Array of
Independent Disks (RAID)
Server connection Failover None
Sever software, such as the
operating system, a
service, or an application

Failover None





Clustering cannot eliminate all possible points of failure. It is designed to
protect availability to data but it cannot protect the data itself. Therefore, it is
still important to have a backup strategy.

Note
Module 1: Introduction to Windows Clustering 9

Scalability
 Enhanced Symmetric Multiprocessing
 Cluster Service
 Network Load Balancing

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Microsoft Windows 2000 Advanced Server provides integrated system
scalability through enhanced symmetric multiprocessing (SMP), in addition to
the two Windows Clustering technologies, Cluster service and Network Load
Balancing service.
Combined with relatively inexpensive computer hardware, Windows 2000
Advanced Server gives organizations powerful and scalable alternatives to more
expensive proprietary solutions.
Enhanced Symmetric Multiprocessing Scalability
SMP is a technology that allows software to use multiple processors on a single
server to improve performance, a concept known as hardware scaling, or

scaling up.
Windows 2000 Advanced Server supports up to 8-way SMP. Improvements in
the implementation of the SMP code allow for improved scaling linearity,
making Windows 2000 Advanced Server an even more powerful platform for
business-critical applications, databases, and Web services. In an SMP system,
several processors share a global memory and I/O subsystem.
At the hardware level, the major drawback to SMP systems is that they
encounter physical limitations in bus and memory speed that are expensive to
overcome. As microprocessor speeds increase, shared memory multiprocessors
become increasingly expensive. There are large cost differences as customers
increase their systems from one processor to 2 to 4 processors, and especially
when implementing more than 8 processors.
Cluster Service
Cluster service is a feature of Windows 2000 Advanced Server that allows a
pair of independent servers, referred to as nodes, to be managed as a single
entity. The objective of Cluster service is to provide high levels of availability
and scalability for applications and data.
Topic Objective
To identify the clustering
technologies that provide
scalability.
Lead-in
Microsoft Windows 2000
Advanced Server provides
integrated system scalability
through enhanced
symmetric multiprocessing,
in addition to the two
Windows Clustering
technologies, Cluster

service and Network Load
Balancing.
10 Module 1: Introduction to Windows Clustering

Network Load Balancing
Network Load Balancing service enables organizations to cluster up to 32
servers running Windows 2000 Advanced Server to evenly distribute incoming
traffic while also monitoring servers and the network. The dual benefits of
simple, incremental scalability combined with high-availability make Network
Load Balancing service ideal for use with business-critical e-commerce,
Internet Service Provider hosting, and Terminal Services applications.
Network Load Balancing service introduces the concept of software scaling; or
scaling out, where system administrators can add capacity to their server farms
by simply plugging in additional Network Load Balancing service-configured
servers as needed.

Module 1: Introduction to Windows Clustering 11

High Availability
 Measuring High Availability
 Cluster Service
 Network Load Balancing

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Windows 2000 Advanced Server provides system services for server clustering
as a standard feature of the product. The objective of clustering is to provide
very high levels of application and data availability.
Availability refers to the percentage of time that a system is available for the
users. Availability is increased by improving reliability and by reducing the

amount of time that a system is down for various reasons, such as planned
maintenance or recovery from failure.
Measuring High Availability
High availability is a measure of the time during which clients can successfully
use a resource, application, or system within design specifications.
Availability is normally expressed as a percentage. For example, a computer
system that is required on a 24x365 basis that is unavailable for 24 hours would
have an availability percentage of 99.62%. To achieve 99.99% availability this
system can only be unavailable for 53 minutes per year. To achieve 99.999%
availability this system can only be unavailable for 5.3 minutes a year.
A computer system with high availability will optimally provide continuous
service without interruptions that are caused by software or hardware failures.
Topic Objective
To identify the clustering
technologies that provide
high availability.
Lead-in
Windows 2000 Advanced
Server provides system
services for server
clustering as a standard
feature of the product. The
objective of clustering is to
provide very high levels of
application and data
availability.
Delivery Tip
Use the example provided
that demonstrates
availability expressed as a

percentage.
Key Point
Ensure that the students
understand that high
availability does not imply
fault tolerance.
12 Module 1: Introduction to Windows Clustering

Comparing High Availability and Fault Tolerance
It is important to note that high availability is not fault tolerance. Fault tolerant
systems, such as those used for air traffic control applications, may be required
to achieve greater than 99.999% availability. This is typically achieved by
adding extensive redundancy to the system hardware, which instantly provide
backup components in the event of primary component failure with no loss of
process or data consistency. An example of a fault tolerant system is RAID
technology. Logical disk data is written to an array of disks with additional
information so that the loss of a single disk can be tolerated without preventing
access to the data. There is no backup component in this case; the data is
dynamically rebuilt from the information on the other drives in the array. You
can remove and replace the failed disk drive with a new one, and the system is
repaired, returning to the initial state before the failure.
A fault tolerant system is designed to guarantee resource availability. A high-
availability system is concerned with maximizing resource availability.
Cluster Service
The use of component hardware provides many advantages, including reduced
purchase cost and greater standardization; these advantages can lead to reduce
maintenance costs. But component hardware, like all hardware, is subject to
periodic failure. Windows 2000 Advanced Server provides for high availability
of these hardware components through the use of clustering.
A cluster is a group of servers that appear to the client as a single entity. The

nodes in a cluster access the same disk drives, so any single server in the cluster
has access to the same set of data and programs. The servers in a cluster act as
backups for each other, if any one server in the cluster stops working, its
workload is automatically moved to another server in the cluster in a process
called failover.
Network Load Balancing
Another way to improve availability is through the use of network load
balancing. Network load balancing is a method where incoming requests for
service are routed to one of several different computers.
Network load balancing is provided by Network Load Balancing services in
Windows 2000 Advanced Server and Microsoft Windows 2000 Datacenter
Server.
Module 1: Introduction to Windows Clustering 13



 Introducing Microsoft Windows 2000 Clustering
ethernet
ethernet
Web
Host 1
Web
Host 1
Web
Host 2
Web
Host 2
Web
Host 3
Web

Host 3
Web
Host 4
Web
Host 4
Network Load Balancing
Component Load Balancing
Network Load Balancing
Component Load Balancing
2-node Cluster Service
2-node Cluster Service
Internet
InternetCustomer
Customer
Database
Messaging
File Shares
Database
Messaging
File Shares

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Windows 2000 Advanced Server provides two clustering technologies that can
be used independently or in combination, Network Load Balancing service,
Component Load Balancing (available with Application Center 2000) and
Cluster service. These technologies provide a complete set of clustered
solutions to choose from depending on your application or service. The
preceding graphic is a graphical representation of both technologies.
 Network Load Balancing. This service load balances incoming Internet

protocol (IP) traffic across clusters of up to 32 hosts. Network Load
Balancing service enhances both the availability and scalability of Internet
server-based programs, such as Web servers, streaming media servers, and
Terminal Services. By acting as the load balancing infrastructure and
providing control information to management applications built on top of
Windows Management Instrumentation (WMI), Network Load Balancing
service can seamlessly integrate into existing Web server farm
infrastructures. Network Load Balancing service will also serve as an ideal
load balancing architecture for use with the Microsoft release of Microsoft
Windows 2000 Application Center 2000 in distributed Web farm
environments.
 Cluster Service. This service is intended primarily to provide failover
support for applications, such as databases, messaging systems, and file and
print services. Cluster service supports 2-node failover clusters in
Windows 2000 Advanced Server and 4-node clusters in Datacenter Server.
Cluster service is ideal for ensuring the availability of critical line-of-
business and other back-end systems, such as Microsoft Exchange Server or
a database running Microsoft SQL Server version 7.0 that is acting as a data
store for an e-commerce Web site.
 Component Load Balancing. This service will be a feature of Microsoft
Application Center 2000. Component Load Balancing distributes workload
across multiple servers running a site’s business logic components.

Topic Objective
To identify the available
Microsoft clustering
technologies.
Lead-in
Windows 2000 Advanced
Server provides two

clustering technologies that
you can use independently
or in combination, Network
Load Balancing, Component
Load Balancing (available
with Application Center
2000), and Cluster service.
Delivery Tip
Explain to the students that
this topic is just an overview
of Windows 2000 clustering.
Don’t spend too much time
on each slide because each
of these technologies will be
covered in detail in the
following modules.
14 Module 1: Introduction to Windows Clustering

Network Load Balancing Service
Network Load Balancing
Network Load Balancing
ethernet
ethernet
Web
Host 1
Web
Host 1
Web
Host 2
Web

Host 2
Web
Host 3
Web
Host 3
Web
Host 4
Web
Host 4
Internet
InternetCustomer
Customer
Database
Messaging
File Shares
Database
Messaging
File Shares

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Windows 2000 Network Load Balancing service provides an integrated
infrastructure for creating a distributed load-balanced environment for your
critical services, such as high-demand Web sites.
Designed for use with a diverse array of applications and services, Network
Load Balancing uses a statistical or manual load-balancing algorithm to
distribute incoming IP requests across a cluster of up to 32 servers.
As the system administrator deploying Network Load Balancing, you will be
able to:
 Scale Web applications by quickly and incrementally adding additional

servers.
 Ensure that your Web sites are always online for your customers. Network
Load Balancing supports load balancing, which reduces poor customer
experience that results from unplanned downtime.

Network Load Balancing service combined with application monitoring
tools that are included in Windows 2000 Resource Kit ensure that your Web
site is always available to customers.

 Scale virtual private network (VPN), Point-to-Point Tunneling Protocol
(PPTP) servers to accommodate every user account with simplified access
to a central IP address.
 Scale streaming media services for performance and scalability.
 Scale Terminal Services to support large user accounts by distributing
connections across multiple servers.

Topic Objective
To introduce the features
and benefits of Network
Load Balancing.
Lead-in
Windows 2000 Network
Load Balancing service
provides an integrated
infrastructure for building
your critical, in-demand Web
sites in a distributed, load-
balanced manner.
Note
Module 1: Introduction to Windows Clustering 15

Component Load Balancing
Network Load
Balancing
Network Load
Balancing
Component Load Balancing (COM+)
Component Load Balancing (COM+)
Clustering Service
Clustering Service
Clients
Clients
IISWeb Server or
other IP-based services
IISWeb Server or
other IP-based services
Application Servers
Application Servers
COM+
Components
COM+
Components
Data Servers
SQL Server, Exchange Server File
Data Servers
SQL Server, Exchange Server File

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Component Load Balancing service, a feature of Microsoft Application Center

2000, load balances different instances of the same COM+ components that are
running on one or more servers. This dynamic load-balancing feature of
Component Load Balancing enables applications that are using COM+
components to be distributed evenly across a group of application servers for
increased reliability and scalability.
In the event of a server failure, Component Load Balancing is notified and re-
routes requests away from the failed node, ensuring continuous availability of
the COM+ components even in the event of multiple system hardware or
software failures.
Component Load Balancing complements both Network Load Balancing
service and Cluster service by acting on the middle-or business tier of a
multitiered clustered network.
The key differences between Network Load Balancing service, Component
Load Balancing, and Cluster service are:
 Network Load Balancing service cannot differentiate between Uniform
Resource Locators (URLs) being sent in Hypertext Transfer Protocol
(HTTP) requests, so it processes all of the requests and all of the clustered
computers identically.
 Component Load Balancing mechanisms offer finer-grained control than
Network Load Balancing service and are able to differentiate between URLs
being sent in HTTP requests and route them in the most efficient manner.
 Cluster service nodes share a disk, which is important for storage services,
such as databases or groupware messaging stores, but provides little benefit
to the majority of COM+ component servers. Unlike Network Load
Balancing and Component Load Balancing, Cluster service cannot work
across more than two nodes in Microsoft Windows 2000 Advanced Server.

Topic Objective
To introduce the features
and benefits of Component

Load Balancing.
Lead-in
Component Load Balancing
service, a feature of
Microsoft Application Center
2000, load balances
different instances of the
same COM+ components
that are running on one or
more servers.
Delivery Tip
Emphasize to the students
that this page is just an
overview of Component
Load Balancing. This
clustering technology will be
covered in more detail in the
Application Center 2000
course.
16 Module 1: Introduction to Windows Clustering

Cluster Service
Network Load Balancing
Network Load Balancing
2-node Cluster Service
2-node Cluster Service
ethernet
ethernet
Web
Host 1

Web
Host 1
Web
Host 2
Web
Host 2
Web
Host 3
Web
Host 3
Web
Host 4
Web
Host 4
Internet
InternetCustomer
Customer
Database
Messaging
File Shares
Database
Messaging
File Shares

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Applications that are central to your organization’s operations include systems
such as databases, messaging servers, enterprise resource planning applications,
and core file and print services. Cluster service in the Windows 2000 operating
system ensures that these critical applications are online when needed by

removing the physical server as a single point of failure.
In the event that a hardware or software failure occurs in either node, Cluster
service migrates the applications currently running on that node to the surviving
node and restarts them. Because Cluster service uses a shared-disk
configuration with common bus architectures, such as small computer system
interface (SCSI) and Fibre Channel, you will not lose any data during a
failover.
As the system administrator deploying Cluster service, you will be able to:
 Reduce unplanned downtime. Downtime caused by hardware or software
failures can result in lost revenue and poor customer experience. Using
Cluster service with a shared-disk solution on critical line-of-business
applications can significantly reduce the amount of application downtime
that unexpected failures cause.
 Deploy upgrades smoothly with rolling upgrade support. Cluster service is
ideally suited for ensuring transparent upgrades of applications without
interrupting your clients. By migrating your applications to one node,
upgrading the first node, and then migrating them back, you can roll out
hardware, software, and even operating system upgrades without taking the
application offline. Cluster service in Windows 2000 supports rolling
operating system upgrades from Microsoft Windows NT
® Server version
4.0, Enterprise Edition clusters that are deployed with Service Pack 4 or
higher.
 Deploy reliable applications. Cluster service is supported by dozens of
cluster-aware applications spanning a wide range of functions and vendors.
Topic Objective
To introduce the concepts of
Cluster service.
Lead-in
Applications that are central

to your organization’s
operations include systems,
such as databases,
messaging servers,
enterprise resource planning
applications, and core file
and print services.
Module 1: Introduction to Windows Clustering 17

 Deploy applications on industry-standard hardware. Cluster service also
allows you to cluster services such as, Dynamic Host Configuration
Protocol (DHCP), Windows Internet Name Service (WINS), and
Distributed File System (DFS).
 Install and configure in less time. Cluster service in Windows 2000 is now
easier to set up and use. With a substantially improved Setup Wizard,
Cluster service setup requires fewer entries and less time to install and
configure. Combined with the improved Cluster Administrator, now a
Microsoft Management Console snap-in, the Cluster service in the
Windows 2000 operating system is redefining how simple building clusters
on standard Intel computer-based hardware can be.

18 Module 1: Introduction to Windows Clustering

Comparing Network Load Balancing to Cluster Service
Which Clustering Technology Should be Used for Your Application?
Which Clustering Technology Should be Used for Your Application?
Which Clustering Technology Should be Used for Your Application?
Cluster
Service
Cluster

Cluster
Service
Service
Networking Load
Balancing
Networking Load
Networking Load
Balancing
Balancing
Benefits
Benefits
Benefits
Web Server
Farm
Web Server
Farm
Terminal
Services
Terminal
Services
File/Print
Servers
File/Print
Servers
Database/
Messaging
Database/
Messaging
Quickly expand your capacity
Minimize site downtime

Quickly expand your capacity
Minimize site downtime
Quickly expand your capacity
Minimize effects of server failures
Quickly expand your capacity
Minimize effects of server failures
Minimize service downtime
Ensure data consistency after failover
Minimize service downtime
Ensure data consistency after failover
Minimize application downtime
Ensure data consistency after failover
Minimize application downtime
Ensure data consistency after failover
E-Commerce
Sites
E-Commerce
Sites
Quickly expand your capacity
Minimize effects of server/app. downtime
Quickly expand your capacity
Minimize effects of server/app. downtime
Scenario
Scenario
Scenario
Technology
Technology
Technology




















*****************************
ILLEGAL FOR NON-TRAINER USE******************************
Windows 2000 Advanced Server and Datacenter Server operating systems
support two clustering technologies that you can use independently or in
combination, Cluster service and Network Load Balancing service.
You can use combined Windows Clustering technologies to create e-commerce
sites that have high scalability and high availability. By deploying Network
Load Balancing service across a Web server farm, and clustering back-end line-
of-business applications, such as databases with Cluster service, you can gain
all of the benefits of near-linear scalability with no server or application-based
single points of failure.
As a system administrator you will need to decide which clustering solution to
implement, Cluster service or Network Load Balancing service. The preceding

graphic demonstrates which clustering technology should be implemented,
depending on the business need, and gives the benefits for each solution.
Topic Objective
To compare Cluster service
and Network Load
Balancing.
Lead-in
Windows 2000 Advanced
Server and Datacenter
Server operating systems
introduce two clustering
technologies that you can
use independently or in
combination.
Module 1: Introduction to Windows Clustering 19

Identifying the Application and Service Environments
 Application Environment
 Services Environment

*****************************
ILLEGAL FOR NON-TRAINER USE******************************
After you have identified the potential points of failure within your system, the
next step is to determine which application and services you can move to the
cluster. Typically, the resources moved to a cluster are those that provide access
to mission-critical data and where loss of access can have a negative impact on
customer experience.
Application Environment
Microsoft developed the Application Specification for Windows 2000 in
cooperation with customers and third-party developers to provide clear, concise

guidelines to help developers create applications that deliver new levels of
reliability and manageability.
Three types of server applications benefit from clustering technologies:
 In the box services of Windows 2000 Advanced Server: These services
include file shares, print queues, Internet/intranet sites managed by
Microsoft Internet Information Services (IIS), Microsoft Terminal Services,
Microsoft Routing and Remote Access VPN Server, Microsoft Message
Queue Server (MSMQ) services, and Component Services, all which are
part of Windows 2000 Advanced Server and Windows 2000 Datacenter
Server.
Topic Objective
To identify resources that
can run within a cluster.
Lead-in
After you have identified the
potential points of failure,
the next step is to determine
which resources, such as
applications, printers, or file
shares, need to be moved to
the cluster.