My Profile
Active Members
TodayLast 7 Days
more...
Awards & Gifts
Online Exams
Fresher Jobs
Our fresher job section is exclusively for fresh graduates! Find jobs for freshers in major Indian
cities including Bangalore, Chennai, Hyderabad, Pune or Kochi
Resources
Find educational articles, blogs, discussion threads and other resources.
Colleges
Find details about any college in India or search for courses.
Advertisements
|
MCSE EXAM MATERIAL 2
Posted Date: 27 Feb 2008 Resource Type: Articles/Knowledge Sharing Category: Computer & Technology
|
Posted By: AFTAB VOHRA Member Level: Bronze
Rating:  Points: 4
|
|
|
EXAM OBJECTIVES
This chapter kicks your preparation for the Directory Services
exam into high gear. Although the basics covered in this chapter
don’t apply to any one specific exam objective, they’re the
fundamental cornerstones on which you’ll build your Active
Directory knowledge and skills.
Active Directory is new for Windows 2000 and has its own unique
set of terminology and components. In this chapter, you’ll master
that terminology, as well as basic Active Directory concepts. With
the basics firmly under your belt, you’ll be ready to meet head-on the
more advanced Active Directory topics covered in later chapters.
Directory Services
EXAM
MATERIAL
Overview of Active
Directory
No overview of Windows 2000 would be complete without a discussion
of Active Directory. In this chapter, I’ll define what Active Directory is
and discuss some of its important features. Next I’ll explain in detail the
structure of Active Directory, including its many components. I’ll go over some
of the practicalities of how Active Directory is implemented. Finally, I’ll introduce
some food for thought when planning for Active Directory on your Windows
2000 network.
CHAPTER
2
43
Chapter Pre-Test
1. Two key features of Active Directory are _____________ and
_____________.
2. What are the fundamental units that make up Active Directory?
3. What is the difference between a domain tree and a forest?
4. What service is Active Directory dependent on?
5. What is the primary purpose of organizational units (OUs)?
44 Part I Introduction to Windows 2000
What Is Active Directory?
Active Directory is the directory service used by Windows 2000. It is a core
new feature of the Windows 2000 operating systems.
A directory service consists of two parts—a centralized, hierarchical
database that contains information about users and resources on a network,
and a service that manages the database and enables users of computers on
the network to access the database. In Windows 2000, the database is called
the Active Directory data store, or sometimes just the directory.The Active
Directory data store contains information about various types of network
objects, including printers, shared folders, user accounts, groups, and
computers.Windows 2000 Server computers that have a copy of the Active
Directory data store, and that run Active Directory are called domain
controllers. In a Windows 2000 domain, a read/write copy of the Active
Directory data store is physically located on each domain controller in the
domain. A domain is a logical grouping of networked computers in which
one or more of the computers has shared resources, such as a shared folder
or printer, and in which all of the computers share a common Active
Directory data store.
The three primary purposes of Active Directory are:
To provide user logon and authentication services
To enable administrators to organize and manage user accounts,
groups, and network resources
To enable authorized users to easily locate network resources,
regardless of where they are located on the network
CROSS-REFERENCE
This chapter is basically a theoretical and planning discussion about
Active Directory. Later chapters in this book address installing Active
Directory (Chapter 7), administering and securing Active Directory
(Chapter 8), and managing and optimizing Active Directory operations
and replication (Chapter 22).
So why is Active Directory so cool? I’ll answer that question in the next
section by discussing some of the features of Active Directory.
Chapter 2 Overview of Active Directory 45
Understanding the Features of
Active Directory
Active Directory is a major step forward for the Windows NT/Windows
2000 operating systems. Just a few of the key features and benefits offered
by Active Directory are:
It provides fully integrated security.
It provides ease of administration by using group policies.
It makes resources easier to locate.
It is scalable to any size network.
It is flexible and extensible.
Because each of these features is fundamental to your understanding of
Active Directory, I’ll discuss them individually in the following sections.
Fully Integrated Security
When I say that Active Directory provides fully integrated security, I am
actually expressing two important concepts. First,Active Directory (working
in conjunction with the Windows 2000 Security subsystem, in which Active
Directory resides) provides network security by managing the logon
and authentication processes. Second, Active Directory (and the Security
subsystem) provides security by controlling access to objects (such as user
accounts, shared folders, and printers) in the directory data store.What makes
this access control so powerful is that it can be defined precisely—not only
on each and every object in the directory data store, but on each separate
property of each object, as well.
Ease of Administration
The logical, hierarchical structure of Active Directory, in conjunction with
group policies, makes for greater ease in administering a Windows 2000
Server network.
You can think of the structure of Active Directory as being like the
hierarchical structure of a file system. When working with a file system,
you can assign a particular user administrative rights to a folder and to all of
that folder’s contents. In Active Directory, you can delegate, to a particular
46 Part I Introduction to Windows 2000
user, administrative rights to a specific part of Active Directory, and to all of
that part’s contents.
The hierarchical structure of Active Directory also lends itself to the
application of group policies. A group policy is a policy that contains rules,
settings, or both that are applied to all users or computers located in a
specific part of Active Directory. For instance, a group policy can be used to
define a set of programs that will appear on the desktops of the users whose
user accounts are located in a particular part of Active Directory.This part
could be a department, a floor in a building, a geographic location, or the
entire company. In Active Directory, you can apply a group policy to your
entire network, or to the largest unit of your organization to which you
want the policy to apply. Administration is easier because you manage
settings that apply to many users by implementing a small number of
policies, rather than by manually configuring settings individually for a large
number of users and computers.
Ease of Locating Resources
Because Active Directory stores information on all network resources in a
centralized data store, it stands to reason that it should be easier for a user
to locate a resource than if this information were distributed throughout
numerous databases on the network.And it really is easier.
Active Directory also enables administrators and users alike to quickly
locate an object anywhere on the network by searching for any known
property of the object. For example, suppose I want to find the e-mail
address of a particular user on my network. I can use the Search menu,My
Network Places, or Active Directory Users and Computers to search for
this user by the user’s first or last name, telephone number, or other known
property of that user account.Assuming I have the appropriate permissions
to view the user’s account information, all information about that user,
including the user’s e-mail address, will be displayed.
Scalability to Any Size Network
The hierarchical structure of Active Directory lends itself to scalability.
Because Active Directory can include multiple domains, it is scalable to any
size network.
Chapter 2 Overview of Active Directory 47
Flexibility and Extensibility
Active Directory can evolve as your business does. It is not a static structure
that, once implemented, can never be changed.
Active Directory is said to be extensible. This means that new classes of
objects can be added, and new attributes can be added to classes of objects
already present.
Now that you have a basic understanding of what Active Directory is and
an awareness of some of its key benefits, it’s time to wade in a little deeper
to the actual structure of Active Directory and its many components.
Understanding the Structure of
Active Directory
To review: Active Directory has a hierarchical, tree-like structure.
Information about network users and resources is stored in the Active
Directory data store, which is a structured, centralized database.A read/write
copy of the Active Directory data store is physically located on each domain
controller in a Windows 2000 domain.This data store is commonly referred
to as the directory.
EXAM TIP
If you don’t have a solid understanding of the structure of Active
Directory and its components, don’t even think of taking the Directory
Services exam! Nailing down these concepts is vitally important to your
success on this exam.
In order to talk in greater depth about the structure of Active Directory,
I need to introduce and define several new terms. Many of these terms are
components of Active Directory, and some of the terms are used to define
relationships between the components. In the following sections I’ll discuss
objects and classes, schema, the global catalog, and the hierarchical structure
of Active Directory, including domains, organizational units, trees, trust
relationships, and forests. I’ll also discuss Active Directory names and
naming conventions, as well as security. When I’m finished, you’ll have a
much better picture of how Active Directory is structured.
48 Part I Introduction to Windows 2000
Objects and Classes
An Active Directory object is a record in the directory that is defined by a
distinct set of attributes. The attributes of an object are the same as the
object’s properties.The terms are synonymous; however, the term properties
is more prevalent throughout the Windows 2000 user interface.
The specific attributes that an object can have are defined by the object’s
class. A class is simply a template that is used to define the attributes of
an object when it is created. A class defines the required and optional
attributes of the objects that are instances of that class. For example, the
Computer class contains a list of the required and optional attributes that
are used when a computer object is created. All computer objects will be
created using the same Computer class definition.
There are many classes of Active Directory objects. Some of the classes are:
Computer
Contact
Group
Organizational Unit
Domain
Printer
User
Shared Folder
Schema
In Active Directory terminology, the schema is a formal definition—a set
of rules, if you wish—of all of the classes of objects and their attributes
that are stored in the directory. The schema governs the structure of the
directory, including how various objects in the directory fit into the directory’s
hierarchical structure.
The schema is what makes Active Directory extensible.As organizations
change, it may be necessary to add or modify object attributes, or even to
create new classes.The use of certain applications, in particular,may require
these kinds of modifications. Microsoft anticipates that application vendors
will provide the means to modify the schema when necessary to support
their application’s specific requirements.
Chapter 2 Overview of Active Directory 49
Windows 2000 Server includes a tool to modify the schema. It is
a Microsoft Management Console (MMC) snap-in that is only available
after installing the Windows 2000 Administration Tools (ADMINPAK)
on a Windows 2000 computer. The name of the snap-in is Active
Directory Schema.
CROSS-REFERENCE
For information on Installing the ADMINPAK, see the sidebar titled
“Installing the ADMINPAK” in Chapter 8.
Because the schema is the heart of Active Directory, it’s important that
it be protected from accidental or unauthorized modification. For this
reason, Microsoft created a special Security Group for Windows 2000
called Schema Admins. Only users with this permission can run programs
that will modify the schema.
Global Catalog
The global catalog is a master, searchable index that contains information
about every object in every domain in a forest. For now, you can think of
a forest as all of the domains that make up a company’s network. Forests will
be covered in more technical detail later in this chapter.
The global catalog, in conjunction with various search tools, is what
enables administrators and users to search for and quickly locate an object,
regardless of where the object is located on the network.
Windows 2000 automatically creates, by default, a global catalog on the
first domain controller that is installed in a forest.You can configure other
domain controllers to maintain a copy of the global catalog, as well.The
global catalog contains a full copy, or replica, of all objects in its host
domain, and a partial replica of all objects in all other domains in the
forest. A partial replica includes the most common properties of every
object, but not all of the properties of every object.
Hierarchical Structure
By now you’ve read the term “hierarchical structure” a zillion times. But
what does it mean, exactly? A hierarchical structure refers to a manner of
organizing a group of interrelated elements in which the elements are
ranked or stacked, one above the other. An example of a hierarchical
50 Part I Introduction to Windows 2000
structure that you are probably familiar with is an organizational chart.
Figure 2-1 shows an organizational chart for ABC Bank.
FIGURE 2-1 Organization chart of ABC Bank
In the organizational chart of this regional bank, the President is at the
top of the chart, and beneath the President is a level consisting of two Vice
Presidents.After the Vice President level is a layer of management staff, and
beneath this layer is a level that represents the numerous bank tellers.The
hierarchical structure typically has at the top a single element, which
branches into lower layers that contain progressively more elements the
farther down you go.
The key building blocks in the Active Directory hierarchical structure
are domains, which are the focus of the next section.
Domains
Domains are the fundamental units that make up Active Directory. As
stated previously, a domain is a logical grouping of networked computers in
which one or more of the computers has shared resources, such as a shared
folder or printer, and in which all of the computers share a common Active
Directory data store that contains user account, resource, security, and
other information.Active Directory consists of one or more domains.
A domain is a natural security boundary in a Windows 2000 network.
Users from other domains cannot pierce this boundary to access shared
resources unless trust relationships are created between the domains to
Vice President
Region #2
Vice President
Region #1
President
Teller1 Teller2 Teller3 Teller4 Teller5 Teller6 Teller7 Teller8
Manager
Branch D
Manager
Branch C
Manager
Branch B
Manager
Branch A
Chapter 2 Overview of Active Directory 51
permit user access. More information about trust relationships is provided
later in this chapter.
A domain can span several geographic locations of a company, or a
domain can be created for every location. Sometimes the needs of the
departments, divisions, or subsidiaries of an organization determine the
number and structure of the domains needed to effectively manage the
organization’s network.
TIP
When possible, I recommend using a single domain, because this greatly
simplifies the administration of your network.
The domains that make up Active Directory usually correspond to the
network’s DNS domains, and typically use the same FQDN naming
convention used by DNS servers. FQDN stands for fully qualified domain
name, and is the naming convention used on the Internet.The format for an
FQDN is server_name.domain_name.root_domain_name. I’ll discuss
names and naming conventions in a bit more detail later in this chapter.
Domains contain objects, and can also contain organizational units,
which are discussed in the next section.
Organizational Units
Organizational units are a type of Active Directory object, and are sometimes
called container objects.They contain objects and other organizational units
from their own domain. Organizational units are often called by their
abbreviated name ( OUs.
An organizational unit is used to organize related objects and other
organizational units in Active Directory in much the same way that a
folder is used to organize related files and other folders in a volume. Also,
the organizational unit is the smallest container component of Active
Directory to which you can delegate administrative authority or assign
group policy. The primary purpose of an organizational unit, then, is the
organization of related objects and other organizational units to simplify
administration.
For example, suppose an administrator wants to delegate network
administration of the Sales department to an assistant administrator.The
administrator decides to group together all of the objects associated with
the Sales department (including users, computers, printers, shared folders,
and groups). Then the administrator creates an organizational unit and
52 Part I Introduction to Windows 2000
places all of the objects associated with the Sales department into this
organizational unit. Completing these steps enables the administrator to
delegate administration for the Sales department by assigning the assistant
administrator the permissions required to administer the organizational
unit and its contents.
Trees
In Active Directory terminology, a domain tree is a hierarchical grouping of
one or more domains that must have a single root domain, and may have
one or more child domains. In a domain tree, the root domain is the
domain at the top (or root) of the tree.
Domains in a domain tree are often spoken of in terms of parent
domains and child domains. A parent domain is any domain that is above
another domain in the domain tree hierarchy.A child domain is any domain
that is below another domain in the tree.A domain can be a parent to a
domain below it and a child to the domain above it. In a multidomain tree,
the root domain is always a parent domain. Figure 2-2 illustrates a domain
tree.Notice that there is only one root domain in the tree, but that the tree
contains more than one child domain.
FIGURE 2-2 A domain tree
Also notice the naming structure used in Figure 2-2. In a domain tree,
the domains that make up the tree have contiguous DNS domain names.
The root domain’s name forms the basis of (and will be a part of) the
root1.com
child1.root1.com child2.root1.com
grandchild.child2.root1.com
Chapter 2 Overview of Active Directory 53
FQDNs of all of the other domains in the tree.A child domain’s FQDN is
created by appending the name of the parent domain to its own NetBIOS
name by using the child_domain.parent_domain.root_domain.
com format. For example, in Figure 2-2, the domain with a NetBIOS
name of child1 appends the name of its parent domain, root1.com., to its
own name, resulting in an FQDN of child1.root1.com. The root
domain in a domain tree also takes its name in this way, by appending the
name of the first-level DNS domain that it is a member of to its own
NetBIOS name. In Figure 2-2, the root domain with a NetBIOS name of
root1 appends the name of the first-level DNS domain, com, to its own
name, resulting in an FQDN of root1.com.
In organizations that require multiple domains, a domain tree enables
any permitted user in any domain in the tree to access shared resources in
any domain in the tree.This user access is made possible by the special trust
relationships that exist between the domains in the tree.
Trust Relationships
To manage the interaction between multiple domains, trust relationships
are necessary. A trust relationship, or trust, is an agreement between two
domains that enables users in one domain to be authenticated by a domain
controller in another domain, and therefore to access shared resources in
the other domain.
The terminology used to discuss trusts is sometimes confusing, so a
good portion of this section is dedicated to explaining and clarifying these
terms. Once you’ve mastered the terminology, trust concepts are much
easier to understand.
Trusting Domain vs. Trusted Domain Two terms are commonly used to
refer to a trust between two domains: trusting domain and trusted domain.
The trusting domain is the domain that has resources to share with user
accounts in the trusted domain. The trusting domain trusts the trusted
domain. The trusted domain is the domain that contains the user accounts
that want to access the shared resources in the trusting domain.The trusted
domain is trusted by the trusting domain.
A trust relationship between two domains is depicted in diagrams by
using an arrow to point from the trusting (resource) domain to the trusted
(user accounts) domain. Figure 2-3 illustrates a trust relationship between
the west.com domain and the east.com domain.The west.com domain
54 Part I Introduction to Windows 2000
is the trusting domain, and the east.com domain is the trusted domain.
Notice that the arrow points toward the domain with the user accounts.
This trust relationship enables users from the east.com domain to
access shared resources located in the west.com domain.
FIGURE 2-3 The west.com domain trusts the east.com domain
Intransitive and Transitive Trusts An intransitive trust is a trust relationship
between two domains that does not extend beyond these two domains to
other domains.An intransitive trust is a one-way trust,meaning that a single
trust relationship exists between the two domains.
Suppose that the a.com domain trusts the b.com domain.Further suppose
that the b.com domain trusts the c.com domain. Figure 2-4 shows these
trust relationships.
FIGURE 2-4 Intransitive trusts
At first glance, it might appear that the user accounts in the c.com
domain are able to access resources in the a.com domain, but this is not
the case.A trust relationship does not exist between the a.com domain and
the c.com domain. Therefore, users in the c.com domain can’t access
resources in the a.com domain.
It is possible to establish a two-way trust relationship between two
domains by creating two, one-way trusts between those domains. In a twoway
trust relationship, two domains trust each other.
A transitive trust is a trust relationship between two Windows 2000
domains in the same domain tree (or forest) that can extend beyond these
two domains to other trusted domains within the same domain tree
(or forest).A transitive trust is always a two-way trust, meaning that both of
a.com b.com c.com
west.com
(resources)
east.com
(user accounts)
Chapter 2 Overview of Active Directory 55
the domains trust each other. By default, all Windows 2000 trusts within a
domain tree (or forest) are transitive trusts.
Transitive trusts are depicted in diagrams by a single line with an arrow
at each end. Figure 2-5 illustrates transitive trusts in a Windows 2000
domain tree.
FIGURE 2-5 Transitive trusts
Notice that in Figure 2-5, transitive trust relationships exist between
each child domain and the parent domain, but that no trust relationship
exists directly between the two child domains. Nonetheless, the transitive
trust relationships make it possible for users in the child1.parent.com
domain to access resources located in both the parent.com domain and
in the child2.parent.com domain. Likewise, users in the
child2.parent.com domain can access resources located in both the
parent.com domain and in the child1.parent.com domain because
of the transitive trusts that connect the three domains.
In Windows 2000 domain trees,Windows 2000 Server automatically
creates two-way, transitive trust relationships between a parent domain and
a child domain when the child domain is created in the domain tree.The
presence of transitive trust relationships between all of the domains in a
Windows 2000 domain tree makes it possible for a user in one domain to
access a shared resource located in any domain in the tree, regardless of
how many domains separate the user and the shared resource.
Windows NT Server 4.0 doesn’t support transitive trusts—it only
supports intransitive trusts.This means that the only type of trust relationship
possible between a Windows 2000 domain and a Windows NT domain is an
intransitive trust.
parent.com
child1.parent.com child2.parent.com
56 Part I Introduction to Windows 2000
Explicit Trusts An explicit trust is a trust that an administrator creates, versus
a trust that is automatically created by Windows 2000.An explicit trust can
be either transitive or intransitive. Explicit trusts are sometimes used when
you need to manage trusts between a Windows 2000 domain and a
Windows NT domain. Explicit trusts are also used in large, multidomain
forests to shorten the path between two domains to shorten the time
required for authentication and logon.
Forests
Earlier in this chapter I said you could think of a forest as being all of the
domains that compose a company’s network. A more technically accurate
definition of a forest is a group of one or more domain trees, linked by
transitive trusts, that shares a common schema and global catalog.
A forest begins with one domain and one domain tree. It’s kind of a
difficult concept to grasp, but when you install Active Directory on the
first domain controller on your network,Windows 2000 creates a domain,
a domain tree, and a forest all at the same time. So, even though you’ve
only installed Active Directory on one computer, you’ve got all of these
big-picture elements created and ready to go. Now the forest can grow as
you add additional domains and domain trees.
IN THE REAL WORLD
There are no tools that enable you to work on two forests at the same
time. If you have a multi-domain network, it can be good to have a root
domain that only contains the administrator account so that changes to
domain structures can be easily made.
Figure 2-6 illustrates a forest that consists of two domain trees. Notice
that this forest contains two root domains, each of which forms the basis
for its own domain tree. Also notice that a single, transitive trust connects
the two domain trees.
Take another look at Figure 2-6, and notice the domain names. By
definition, the domains in a domain tree have contiguous DNS domain
names. In this example, rootA.com is contained in the name of every
domain in its tree. Likewise, rootB.org is contained in the name of every
domain in its tree. However, that’s as far as it goes.The two domain trees
themselves do not have contiguous DNS domain names, even though they
have been joined together in a forest.
Chapter 2 Overview of Active Directory 57
FIGURE 2-6 A forest
A forest takes its name from the root tree, which is the first tree created
in the forest.
Names and Naming Conventions
Names are of critical importance in Active Directory. In this section, I’ll
explain the types of names and naming conventions used by Active Directory.
Within Active Directory, each object has a name.When you create an
object in Active Directory, such as a user or a computer, you assign the
object a name.This name must be unique within the domain—you can’t
assign an object the same name as any other object (regardless of its type)
in that domain. If you have a user named AlanC, for example, you can’t
create a computer account in the domain that is also named AlanC.
For more information on developing names for domains, organizational
units, users, groups, and computers, see the section titled “Planning
Naming Conventions” later in this chapter.
At the same time that you create an object, not only do you assign a
name to the object, but Active Directory also assigns identifiers to the
object. Active Directory assigns every object a globally unique identifier
(GUID), and assigns many objects a security identifier (SID). A GUID is
typically a 32-digit hexadecimal number that uniquely identifies an object
within Active Directory. A SID is a unique number created by the
Windows 2000 Security subsystem that is assigned only to security principal
objects (users, groups, and computers) when they are created.Windows
2000 uses SIDs to grant or deny a security principal object access to other
objects and network resources.
Active Directory uses a hierarchical naming convention that is based on
Lightweight Directory Access Protocol (LDAP) and DNS standards.
rootA.com
childA.rootA.com childB.rootA.com
rootB.org
childA.rootB.org childB.rootB.org
58 Part I Introduction to Windows 2000
Objects in Active Directory can be referenced by using one of three Active
Directory name types:
Relative distinguished name (RDN)
Distinguished name (DN)
User principal name (UPN)
A relative distinguished name (RDN) is the name that is assigned to the
object by the administrator when the object is created. For example, when
I create a user named AlanC, the RDN of that user is AlanC.The RDN
only identifies an object—it doesn’t identify the object’s location within
Active Directory. The RDN is the simplest of the three Active Directory
name types, and is sometimes called the common name of the object.
A distinguished name (DN) consists of an object’s RDN, plus the object’s
location in Active Directory. The DN supplies the complete path to the
object. An object’s DN includes its RDN, the name of the organizational
unit(s) that contains the object (if any), and the FQDN of the domain. For
example, suppose that I create a user named AlanC in an organizational
unit called US in a domain named Exportsinc.com.The DN of this user
would be:
AlanC@US.Exportsinc.com
A user principal name (UPN) is a shortened version of the DN that is
typically used for logon and e-mail purposes.A UPN consists of the RDN
plus the FQDN of the domain. Using my previous example, the UPN for
the user named AlanC would be:
AlanC@Exportsinc.com
Another way you can think of a UPN is as a DN stripped of all organizational
unit references.
Security
As mentioned previously, Active Directory resides in the Windows 2000
Security subsystem.Together,Active Directory and the Security subsystem
protect Active Directory against unauthorized access. The Active
Directory/Security subsystem team uses access control lists (ACLs) to
determine who can access (and/or modify) an object. An ACL is a list of
SIDs and the associated access privileges assigned to each SID. Each object
and network resource has an ACL associated with it.
Chapter 2 Overview of Active Directory 59
Another security feature of Active Directory is delegation. As stated
previously, you can delegate administrative privileges for a container object
(such as an organizational unit or domain) and all of its contents to a user
or group.This feature enables you to distribute administrative tasks among
several employees without having to give each of these employees administrative
privileges to the entire network.
Understanding How Active Directory
Is Implemented
My goal in this section is to explain how Active Directory is actually
implemented on a network. I know you’re up to your eyeballs in theory
right now, but hang in there just a little longer and I think you’ll begin to
understand how the pieces fit together.
In the next sections I’ll talk a bit about what happens when Active
Directory is installed, how the global catalog server works to provide
Active Directory information about every object in every domain in a
forest, and what Active Directory replication is and how it’s implemented
on a Windows 2000 network. I’ll also introduce the concept of flexible
single master operations (FSMO) and discuss the various roles that domain
controllers can play. Finally, I’ll explain how Active Directory is dependent
on DNS and cover some of the limitations that non-Windows 2000 client
computers have in terms of using the features of Active Directory.
Installing Active Directory
You have to install Active Directory on all Windows 2000 Server computers
that you want to function as domain controllers for a domain—Active
Directory is not installed by default.When you first install Active Directory
in a domain, Windows 2000 performs two tasks: First, it promotes the
computer on which you’re installing Active Directory to a domain
controller; and second, it creates the Active Directory data store (for that
domain) on the newly created domain controller.When you install Active
Directory on an additional Windows 2000 Server computer in that domain,
Windows 2000 promotes that computer to a domain controller and copies a
read/write replica of the Active Directory data store from one of the other
existing domain controllers in the domain. A Windows 2000 Server
60 Part I Introduction to Windows 2000
computer can be a domain controller for only one domain; however, a
domain can have multiple domain controllers..
Active Directory is installed by using the Active Directory Installation
Wizard that is available in the Configure Your Server tool in the
Administrative Tools folder.You can also start this wizard by selecting
Start?Run, typing Dcpromo.exe, and pressing Enter.
CROSS-REFERENCE
I’ll give detailed instructions for installing Active Directory in Chapter 7.
However, I strongly recommend that you read the rest of this chapter,
including the “Planning for Active Directory on Your Network” section,
before you install Active Directory.
Global Catalog Server
A global catalog server is a domain controller that has an additional duty—it
maintains a global catalog.You may recall that a global catalog is a master,
searchable database that contains information about every object in every
domain in a forest. The global catalog contains a complete replica of all
objects in Active Directory for its host domain, and contains a partial replica
of all objects in Active Directory for every other domain in the forest.
A global catalog server performs two important functions:
Provides group membership information during logon and
authentication
Helps users locate resources in Active Directory
I’ll discuss both of these functions in the next section. A global catalog
server provides universal group membership information during a user’s
logon and authentication process. (A universal group is one that can contain
users and other groups from any domain in the forest.) Determining group
membership information is a critical part of the logon process, because the
groups a user is a member of help determine that user’s rights and
permissions. The global catalog server provides this group membership
information in a highly efficient manner because a global catalog server can
respond to the request, instead of having to query a domain controller from
each domain in the forest. If a global catalog server is not available, the user
will not be able to log on to the domain unless that user is a member of the
Domain Admins group.
Chapter 2 Overview of Active Directory 61
A global catalog server helps users locate Active Directory objects,
regardless of which domain in the forest contains the object. Users can
browse the global catalog for available services and resources. In addition,
because the global catalog contains information on every object in every
domain in the forest, users can search the global catalog for a specific object
or resource. For example, suppose I want to locate the e-mail address of
another employee in my company. I know the employee’s name, so I can
query the global catalog by the employee’s name, and, assuming I have the
appropriate permissions, view pertinent data about the employee I’m
searching for, including his or her e-mail address.
A global catalog server, then, provides unified Active Directory information
across all domains in the forest; whereas domain controllers only contain
information about objects in their own domain. Global catalog servers
are a critical part of a multidomain network.
Each domain maintains its own global catalog server. And, by default,
there is only one global catalog server in each domain. Normally this is a
good idea, but occasionally there may be a valid reason for having more
than one global catalog server in a domain.
CROSS-REFERENCE
For more information on optimizing global catalog servers, see Chapter 22.
By default, the first domain controller established in a domain serves as the
global catalog server.To move the global catalog to a different server or to
add an additional global catalog server, you can use the Active Directory Sites
and Services tool. This tool, which is contained in the Administrative
Tools folder, is available after Active Directory is installed.
Replication
The term replication, as applied to Active Directory, refers to the process of
copying information and information updates from the Active Directory data
store on one domain controller to other domain controllers. The purpose
of replication is to synchronize Active Directory data among the domain
controllers in the domain and forest. Several types of Active Directory
information get replicated:
The schema: The schema is replicated to all domain controllers in
the forest.
62 Part I Introduction to Windows 2000
Configuration data: This data, which includes high-level
forest/tree/domain structure, trust, and configuration information,
is replicated to all domain controllers in the forest.
Domain data: This complete, detailed information about every
object in the domain is replicated only to the domain controllers
within this domain.
Replication of Active Directory data is usually partial, meaning that only
updated information (versus a complete copy of the Active Directory data
store) is copied from one domain controller to other domain controllers.
Typically the only time a complete replication is performed is when you
install a new domain controller on the network.
Windows 2000 automatically performs replication.Many administrators of
small to medium-sized networks will never have to configure replication.
That said, replication uses a fair amount of network bandwidth, so sometimes
it is beneficial to manage when and how replication takes place, particularly
over slow WAN links.You can manage Active Directory replication by using
sites, and by using Active Directory Sites and Services.
Sites
A site consists of one or more TCP/IP subnets, which are specified by an
administrator. Additionally, if a site contains more than one subnet, the
subnets should be connected by high-speed, reliable links. Sites do not
correspond to domains:You can have two or more sites within a single
domain, or you can have multiple domains in a single site.A site is solely a
grouping based on IP addresses. Figure 2-7 shows two sites connected by a
slow WAN link.
FIGURE 2-7 Two sites
256 Kbps WAN Link
Seattle
• Consists of specified TCP/IP
subnets ranging from
10.0.0.1 to 10.127.255.254
• All subnets connected by 100
Mbps Ethernet links
• Consists of specified TCP/IP
subnets ranging from
10.128.0.1 to 10.255.255.254
• All subnets connected by 10
Mbps Ethernet links
Indianapolis
Chapter 2 Overview of Active Directory 63
The purpose of sites is to enable servers that regularly copy data to other
servers (such as Active Directory replication data) to distinguish between
servers in their own site (which are connected by high-speed links) and
servers in another site (which are connected by slower-speed WAN links).
Replication between domain controllers in the same site is fast, and
typically administrators can permit Windows 2000 to automatically
perform this task. Replication between a domain controller in one site and
domain controllers in other sites is slower (because it takes place over a
slow WAN link) and often should be scheduled by the administrator so
that use of network bandwidth for replication is minimized during the
network’s peak-activity hours.
Sites and Active Directory replication can be configured by using Active
Directory Sites and Services.
CROSS-REFERENCE
For detailed information on how to manage and optimize Active Directory
replication, including how to use sites, see Chapter 22.
Flexible Single Master Operations
When Microsoft designed Windows 2000, its goal was to have every
domain controller equal; instead of having a primary domain controller
(PDC) and backup domain controllers (BDCs) like Windows NT 4.0 had,
Microsoft wanted to have one class of domain controller that could
perform every domain controller–related task. The advantages of this
design would be the distribution of server load; the elimination of the need
to connect (sometimes across multiple WAN links) to a specific server for
the creation of users, change of passwords, and so on; and the elimination
of the need to have one server always available to all users.
TIP
When more than one domain controller is able to perform a specific task,
that task is referred to as a multiple master operation. When only one
domain controller can perform a specific task, that task is called a single
master operation.
So, things were going along pretty smoothly, and then, when Microsoft
implemented Active Directory and all of its associated processes, it
discovered that a purely multiple master design just wasn’t going to work
for Windows 2000. Although most domain controller–related tasks can be
64 Part I Introduction to Windows 2000
performed by any domain controller, a few critical tasks had to be limited to
one domain controller in a domain, or to one domain controller in a forest,
or utter mayhem and havoc would result.What we’ve ended up with, then,
is a largely multiple master design, with some restricted single master operations.
These operations are called flexible single master operations (FSMO).
The term flexible refers to the fact that an administrator can choose which
domain controller will perform each restricted single master operation.
There are five different types of flexible single master operations roles
(sometimes called master roles) that a domain controller can perform:
schema master, domain naming master, PDC emulator, relative ID master,
and infrastructure master. Each of these roles defines a specific set of
flexible single master operations that only the domain controller assigned
to that role can perform.
When you first install Active Directory on the first domain controller in
the forest, that domain controller automatically assumes all five of the
flexible single master operations roles. As you add domain controllers, you
can manually reassign or transfer these master roles to other domain
controllers as needed.
Schema Master
The schema master is the only domain controller that can make changes to
the schema.When you, as an administrator, use an application to change
the schema, you don’t necessarily need to sit down at the schema master to
run this application, nor do you need to know which computer is
functioning as the schema master.Windows 2000 seamlessly connects the
application to the schema master (across the network) in order to make the
desired change.
Because the schema is identical throughout the forest, there can be only
one schema master in a forest.
If the computer that is functioning as the schema master is not available
when you want to make a change to the schema, you won’t be able to
change the schema until the schema master becomes available, or until you
assign the schema master role to a different domain controller in the forest.
Domain Naming Master
The domain naming master is the only domain controller that can add or
remove domains to or from the forest. The primary reason for isolating
these tasks is to ensure that when a domain is created, its name is unique
within the forest.
Chapter 2 Overview of Active Directory 65
When you create a new domain in an existing forest by installing Active
Directory on the first domain controller in the new domain, the new
domain controller contacts the domain naming master to verify that the
new domain name is not already in use in the forest, and then, once the
domain naming master determines that the domain name is unique, it adds
the new domain to the forest.
There can be only one domain naming master in a forest.
If the computer that is functioning as the domain naming master is not
available when you want to add or remove a domain, you won’t be able to
add or remove the domain until the domain naming master becomes
available, or until you assign the domain naming master role to a different
domain controller in the forest.
PDC Emulator
The PDC emulator performs one of two different roles, depending on how
Active Directory is implemented on your network.
When Active Directory is configured to interact with Windows NT 4.0
backup domain controllers (BDCs), or to interact with computers that
don’t have Windows 2000 Directory Service Client software, Active
Directory is said to be operating in mixed-mode.When Active Directory
operates in mixed-mode, the PDC emulator acts like a Windows NT
primary domain controller (PDC). In this situation, the PDC emulator
synchronizes user account information (such as user names and passwords)
with the existing Windows NT 4.0 BDCs. In addition, when administrators
or users of computers that don’t run Windows 2000 client software need
to make a user account change, that computer must contact the PDC
emulator to make the desired change.
When Active Directory is configured to interact only with Windows 2000
domain controllers and computers that run Windows 2000 Directory Service
Client software, Active Directory is said to be operating in native-mode.
When Active Directory operates in native-mode, the PDC emulator receives
password changes more quickly than other domain controllers in the domain.
When this occurs, the PDC emulator is said to receive preferential treatment
for replication of password changes. Because of this preferential treatment, the
PDC emulator is the domain controller that is most likely to have the most
current version of a user’s password.Therefore, if another domain controller
fails to authenticate a user due to an apparently incorrect password, it will
forward the user’s authentication request to the PDC emulator, and then
66 Part I Introduction to Windows 2000
convey the PDC emulator’s authentication response (either accept or deny) to
the user.
There can be only one PDC emulator in each domain in a forest.
If the computer that is functioning as the PDC emulator is not available
when you want to perform tasks that require it, you won’t be able to
perform these tasks until the PDC emulator becomes available, or until you
assign the PDC emulator role to a different domain controller in the domain.
Relative ID Master
You may recall that when security principal objects (users, groups, and
computers) are created, Active Directory assigns each of these objects a
security identifier, or SID.An SID consists of two parts: a domain SID and
a relative ID.The domain SID identifies the domain in which the object is
created, and is the same for all objects created in the domain.The relative
ID identifies the object in the domain, and is unique for each object
created in the domain.
The relative ID master (sometimes called the RID master) is the domain
controller in the domain that assigns a range of relative IDs to each domain
controller in the domain for use in creating SIDs. Because of this assignment
by the relative ID master, the potential for domain controllers issuing
duplicate SIDs to newly created security principal objects is eliminated.
There can be only one relative ID master in each domain in a forest.
If the computer that is functioning as the relative ID master is not
available when a domain controller exhausts its assigned range of relative
IDs, that domain controller won’t be able to issue SIDs until the relative ID
master becomes available, or until you assign the relative ID master role to
a different domain controller in the domain.
Infrastructure Master
The infrastructure master is the domain controller in the domain that updates
group membership information when group members (who are users
from other domains) are renamed or moved. For example, say that you
have a group named Accounting in your domain. PatL, a user from another
domain, is a member of the Accounting group. PatL recently changed her
name due to marriage, so you change her user name to PatC. The
infrastructure master is responsible for updating the Accounting group
membership information to reflect the change in the user’s name. (I know
the description of this master role sounds bizarre, but this is really how
it works.)
Chapter 2 Overview of Active Directory 67
There can be only one infrastructure master in each domain in a forest.
If the computer that is functioning as the infrastructure master is not
available when you perform tasks that require it, group membership
information won’t be correctly updated until the infrastructure master
becomes available, or until you assign the infrastructure master role to a
different domain controller in the domain.
CROSS-REFERENCE
For information on optimizing flexible single master operations and
master roles, see Chapter 22.
The Domain Name System
On the Internet, host names are stored in various domains and subdomains
that form a hierarchical tree structure called the Domain Name System
(DNS). A Windows 2000 computer that has the Domain Name System
service installed on it is referred to as a DNS server. (More detail about the
DNS service appears later in this chapter.)
Active Directory uses the same hierarchical naming conventions as DNS.
Because of this, client computers use DNS servers to locate Active Directory
domain controllers. Without DNS, Active Directory couldn’t function,
because client computers wouldn’t be able to locate domain controllers.
The Domain Name System (DNS) service that ships with Windows
2000 supports the dynamic update of the DNS database—this feature was
not supported by the Microsoft DNS Server service in Windows NT Server
4.0. When I use the phrase “dynamic update of the DNS database,” what
I mean is that client computers and servers can dynamically register their
host names and IP addresses with the DNS server, without administrator
intervention. Previous versions of DNS required the administrator to
manually enter host names and their associated IP addresses for each
computer on the network.
You’re probably getting the idea about now that Active Directory is
dependent on DNS. In fact, Active Directory requires a DNS server that
supports SRV (service) resource records (RFC 2052). If you have an
existing DNS server on your network that meets these requirements, you
can use this DNS server for your Windows 2000 network. If you don’t
have an existing DNS server on your network (or have a DNS server but
it doesn’t meet these requirements), you can either install a Windows 2000
stand-alone server with the Domain Name System (DNS) service installed
68 Part I Introduction to Windows 2000
on it to function as your network’s DNS server; or you can choose to
install the Domain Name System (DNS) service during the installation of
Active Directory on the first domain controller on your Windows 2000
network, and thereby make that domain controller into your Windows
2000 network’s DNS server.
CROSS-REFERENCE
I’ll cover DNS in more depth, including detailed installation instructions,
in Chapter 7.
Using Clients with Active Directory
When Active Directory is implemented on a Windows 2000 network,
different types of client computers have varying degrees of functionality
with Active Directory.
In the most ideal of situations (from Microsoft’s point of view, at least),
you would have a pure Windows 2000 network, consisting entirely of
Windows 2000 Server computers and client computers that run Windows
2000 Professional. In this scenario, all of the client computers (assuming
their users had the appropriate permissions) would be able to log on to the
domain; locate shared printers, files, and folders; and browse and search
Active Directory for available resources. In addition, the administrator
could efficiently manage the users and computers of this network by using
group policies—one of the key benefits of Active Directory.
However, in reality, a lot of network administrators aren’t going to
upgrade all of their computers to Windows 2000 (at least not right away,
and maybe never). In this case,Active Directory is configured to operate in
mixed-mode, and client computers retain the same functionality they had
in a Windows NT 4.0 network, but typically they aren’t able to use many
of the new features of Active Directory.
For example, when functioning as client computers on a Windows 2000
network, Windows NT 4.0 computers (both Server and Workstation)
retain the same functionality they have always had.Assuming the users have
the necessary permissions, users are still able to log on to the domain;
access shared printers, files, and folders; and browse the network (but not
browse or search Active Directory). There is currently no Directory
Service Client software available for Windows NT 4.0 computers.
Windows NT 4.0 computers on a Windows 2000 network can’t be
Chapter 2 Overview of Active Directory 69
managed by using group policy—they can only be managed by using
Windows NT 4.0 system policy.
Windows 98 and Windows 95 computers retain the functionality they
had when they composed a Windows NT 4.0 network, and in addition,
when Directory Service Client software is installed (assuming the users
have the appropriate permissions), users are able to browse and search
Active Directory. However,Windows 98 and Windows 95 computers can’t
be managed by using group policy—they can only be managed by using
Windows 98 or Windows 95 system policies.
What the whole client issue boils down to is this:
If your network consists of only Windows 2000 Server and Windows
2000 Professional computers, all computers on the network can
use the features of Active Directory and can be managed by using
group policy.
If your network consists of only Windows 2000 Server,Windows
2000 Professional,Windows 98, and Windows 95 computers, all
computers on the network can use the features of Active Directory
(provided Directory Service Client software is installed on the
Windows 98 and Windows 95 computers). However, only the
Windows 2000 computers can be managed by using group
policy—the Windows 98 and Windows 95 computers can’t.
If your Windows 2000 network includes Windows NT 4.0
computers, the Windows NT 4.0 computers can’t use the features
of Active Directory—unless you upgrade them to Windows 2000.
In addition, only the Windows 2000 computers can be managed
by using group policy—the Windows NT 4.0 computers can’t.
By not providing Directory Service Client software for Windows NT
4.0 computers, Microsoft appears to be saying that customers must
upgrade Windows NT 4.0 computers to Windows 2000 if they want to
take advantage of the benefits of Active Directory.
CROSS-REFERENCE
I’ll cover installing the Directory Service Client in Chapter 4.
70 Part I Introduction to Windows 2000
Planning for Active Directory on
Your Network
Like any other major network change, implementing Active Directory on
your network deserves an appropriate amount of careful planning to
ensure a smooth transition.
There are several things you should consider when planning the implementation
of Active Directory. Three important matters that should be
carefully thought out are your domain design, the naming conventions
you’ll use, and how existing client computers will fit into your overall plan.
I’ll cover some planning considerations for each of these elements in the
sections that follow.
Planning a Domain Design
Planning an Active Directory domain design for your network is one of
the most important tasks you may ever be called on to perform. The
domain design is where it all starts—and ultimately, the decisions you
make here can have wide-ranging effects on your company.
When you plan a domain design, you not only determine the domain
structure you want to use, but also you plan an organizational unit (OU)
structure and you plan for the upgrade of any previous domains. The
following sections discuss these topics.
Planning a Domain Structure
In most situations, in terms of ease of network administration, the best
possible Active Directory domain structure for a Windows 2000 network is
a single domain.When a single domain is used, nearly every administrative
task is simpler than when multiple domains are used, and there are none of
the additional administrative tasks associated with a multidomain structure,
such as managing trust relationships. In addition, when a single domain is
used, users can locate resources easier and more consistently because every
object’s attribute is present in the domain’s global catalog, while not every
attribute of every object is replicated to every global catalog server in a
multidomain design.
A speaker at a conference I recently attended stated that anyone who
didn’t choose a single domain structure for the implementation of Windows
2000 Active Directory was stupid.While for the most part I agree with the
Chapter 2 Overview of Active Directory 71
sentiment behind the statement, there are situations where a multiple
domain structure deserves some consideration. You might find that the
benefits of using multiple domains outweigh the disadvantages when:
The management structure of your organization is very decentralized,
and no one individual has control over the company’s network
design and implementation. Instead of a single individual managing
the company’s global network plan, many individuals in different
departments plan for their own department, division, or location.
A parent company has several subsidiaries and does not want
to integrate the network or the management structure of the
subsidiaries into the parent company’s structure.
Your organization has multiple divisions and it is likely that
one or more of the divisions may be sold or spun off as an
independent entity.
You have multiple locations and some of the locations don’t have
a reliable WAN link to corporate headquarters.You must have a
reliable WAN link to keep Active Directory updated when a
single domain spans multiple locations.
Your existing network uses a multiple domain structure, and you
don’t have the necessary time, money, or manpower to change the
domain structure at this time.
Your company is so large that you anticipate that a single domain
in Active Directory would contain millions of objects.
TIP
Windows NT 4.0 had a functional limitation of about 40,000 objects
(or 40MB of disk space) in a single domain. So, while there is still some
limitation, Windows 2000 can accommodate many more objects in a
single domain than Windows NT 4.0 could.
Planning an Organizational Unit Structure
Organizational units (OUs), in my opinion, are meant to serve exactly one
purpose: to make network administration easier. They’re not meant to
mimic the company’s organizational chart (although if you administer your
network in this manner, then it makes perfect sense to structure OUs in
the same way). Organizational units make it easy to delegate authority to
assistant administrators and to administer the network in manageable
72 Part I Introduction to Windows 2000
chunks. Therefore, organizational unit structure should be designed with
delegation in mind.
If you administer your network floor-by-floor and building-by-building,
then you would probably want your OUs to consists of floors and buildings.
If you administer your network by departments and location, you might
want to use OUs that consist of departments and geographic locations.
I think you get the picture.
CAUTION
Naming an OU after a geographic location can be a risky thing — just
think of all of the cities and countries that have changed their names in
the past 10 years.
One last note on OUs. OUs are not security principal objects. Because
of this, OUs can’t be used in the same way that a group can be used to
assign rights and permissions to users or groups contained within the OU.
If you want to assign rights and permissions to multiple users, use a group.
Planning the Upgrade of Previous Domains
If your network is not a brand new Windows 2000 network, you may need
to plan for upgrading previously existing Windows NT 4.0 domains to
Windows 2000 domains.
In many cases, you may want to move from a single or multiple master
domain model (that consists of several domains) to a single domain model.
If this is your situation, you must decide where all of the objects from each
existing domain will be placed in the new Windows 2000 domain.
For example, you might want to place all of the users, groups, and computers
from each existing domain into OUs in the new Windows 2000
domain that correspond to the previous Windows NT 4.0 domains.
Or, since you’re taking the time to redesign your network anyway, you
might want to totally reorganize your network—one user, computer, and
group at a time. Sometimes this is the best way to go, particularly if the
network has been added to, patched, and otherwise monkeyed with by
numerous administrators over the years.
Planning Naming Conventions
When planning the naming conventions you’ll use on your Windows 2000
network, there are several decisions to be made.You’ll need to plan how
Chapter 2 Overview of Active Directory 73
domains, organizational units (OUs), users, groups, and computers will
be named.
TIP
When planning and implementing any naming convention, I recommend
that you attempt to keep all names intuitive, short, and simple. This will
make everyone’s life (especially yours) much easier.
Naming Domains
First of all, you’ll need to name your Active Directory domain (or domains).
When you do this, you’ll need to consider if the name you assign to your
domain will be the same when accessed by users on your company’s
intranet as it will when accessed by external users over the Internet.
As was previously stated, Active Directory domains use DNS domain
names.Keep in mind that the maximum length of a fully qualified domain
name (FQDN), including periods and all extensions, is 63 characters.
Allowed characters include uppercase letters (A–Z), lowercase letters (a–z),
numbers (0–9), and the hyphen (-).
If your domain name will be the same for both internal and external users,
I recommend choosing a domain name that is as close to your company’s
name as possible so it will be easily recognized and located by Internet users
and by Internet search engines.
If you decide you want to further isolate your company’s private
intranet behind a firewall, you might choose to use one domain name for
internal use, and use a different domain name for external users on the
Internet. In this case, the domain name for internal use can be anything
you want, and the domain name that external users on the Internet use
should be as close to your company’s name as possible.
Before you choose your domain name, you should consider using an
Internet search engine to determine if the name you want to use is already
registered to someone else.Currently, DNS domain names are managed by
InterNIC (which stands for Internet Network Information Center).
Finally, if your network is connected to the Internet, you must register
your company’s DNS domain name (as accessed by users over the Internet)
with the appropriate naming authority (InterNIC in the United States).
Your Internet service provider will usually perform this task for you.
74 Part I Introduction to Windows 2000
Naming Organizational Units
After you’ve named your domain, you’ll want to plan how to name your
organizational units (OUs).The key point about working with OUs is that
they should represent the portion of the organization that is being
managed. For example, if you use OUs to manage the users, groups, and
computers in your company that are located on a particular floor or in a
certain building, the name of the OU should readily identify the floor or
building being managed.
Depending on the number of OUs in your organization, you may not
need a formal naming scheme for OUs. Instead, you can choose intuitive
names for the OUs that represent the particular grouping of objects
they contain.
Naming Users, Groups, and Computers
If you have more than a few people in your organization, you’ll need to
plan a naming convention to use for users, groups, and computers.
When you create user, group, and computer accounts, keep in mind a
few rules for these names:
Length: User logon names can be from one to 20 characters long.
Computer names should be limited to 15 characters in length for
backward compatibility with NetBIOS applications and older
client operating systems, such as Windows 95,Windows 98,
Windows NT Workstation 4.0, and so on.
Uniqueness: Names created in a domain must be unique within
the domain. If you have a user named AlanC, for example, you
can’t create a computer account in the domain that is also
named AlanC.
Unacceptable characters: The following characters can’t be used
in user and group account names:
“ / \ [ ] : ; | = , + * ? < >
In addition, a user or group account name can’t consist entirely of
spaces or periods.
Acceptable characters: As stated previously, computer names can
consist of only the following allowed characters: uppercase letters
(A–Z), lowercase letters (a–z), numbers (0–9), and the hyphen (-).
Chapter 2 Overview of Active Directory 75
There are probably as many naming schemes for users, groups, and
computers as there are network administrators. Often, the overall length of
a name is limited to eight characters so that the name is compatible with
MS-DOS directory name limitations. This eight-character limitation is
common, but certainly not mandatory, especially on most of today’s
networks.
A few common naming conventions for user names include:
The first seven letters of the user’s first name plus the first letter of
the user’s last name
The first letter of the user’s first name plus the first seven letters of
the user’s last name
The user’s initials plus the last four digits of the user’s employee
number
Various hybrid combinations of the preceding schemes
Finally, you’ll need to come up with a way to handle exceptions. It’s
quite common, for example, for two users to have the same first name and
last initial, such as Mike Sinclair and Mike Saunders. If you choose to adopt
the first naming scheme in the preceding list, you will need to have a way
to resolve these potentially duplicate user names.You could resolve the
problem by assigning Mike Sinclair the user account name of MikeS
(assuming he was hired first), and assigning Mike Saunders the user
account name of MikeSa.
Planning for Clients
Earlier in this chapter, I discussed the limitations that non-Windows 2000
client computers have in terms of their ability to use many of the new
features of Active Directory.Well, it’s important to consider how your
existing client computers will fit into your overall Windows 2000 Active
Directory implementation plan.
As I see it, you’ve basically got four options:
To achieve optimum functionality of client computers with Active
Directory, you can upgrade all of the client computers on your
network to Windows 2000 Professional.
76 Part I Introduction to Windows 2000
To achieve moderate functionality of client computers with
Active Directory, you can install the Directory Service Client
on all Windows 98 and Windows 95 computers and upgrade all
other client computers (including Windows NT 4.0 computers)
to Windows 2000 Professional.
To achieve limited functionality of client computers with
Active Directory, you can install the Directory Service Client
on all Windows 98 and Windows 95 computers, and upgrade all
other client computers (including Windows NT 4.0 computers)
to Windows 2000 Professional only as they are replaced.
To achieve minimal functionality of client computers with Active
Directory, you can do nothing to existing client computers now,
and later upgrade client computers to Windows 2000 Professional
as they are replaced.
The choice you make will depend on many factors, including management
desires, the amount of funding you have to implement the project,
the amount of support manpower available, and how important full Active
Directory functionality is to you and your organization.
This chapter introduced several key Active Directory terms and concepts:
Active Directory is the directory service used by Windows 2000. In Windows
2000, the directory service database is called the Active Directory data store.
A read/write copy of the Active Directory data store is physically located on
each domain controller in a Windows 2000 domain.
Active Directory has many key features. It provides fully integrated security,
provides ease of administration by using group policies, makes resources
easier to locate, is scalable to any size network, and is flexible and extensible.
Numerous Active Directory terms and concepts were defined and discussed in
this chapter:
Object: A record in the directory that is defined by a particular set
of attributes
KEY POINT SUMMARY
Chapter 2 Overview of Active Directory 77
Class: A template used to create a specific type of object
Schema: A formal definition of all of the classes of objects and their
attributes stored in the directory
Global catalog: A master, searchable index that contains information
about every object in every domain in a forest
Hierarchical structure: A manner of organizing a group of interrelated
elements in which the elements are ranked or stacked, one above the other
Domain: A logical grouping of networked computers in which one or
more of the computers has shared resources and in which all of the
computers share a common Active Directory data store
Organizational unit (OU): A type of Active Directory object, sometimes
called a container object, that can contain objects and other organizational
units
Domain tree: A hierarchical grouping of one or more domains that must
have a single root domain, and may have one or more child domains
Trust relationship or trust: An agreement between two domains that
enables users in one domain to be authenticated by a domain controller
in another domain, and therefore to access shared resources in the
other domain
Forest: A group of one or more domain trees, linked by transitive trusts,
that shares a common schema and global catalog
Global catalog server: A domain controller that maintains a global catalog
Replication: The process of copying information and information updates
from the Active Directory data store on one domain controller to other
domain controllers
Site: One or more TCP/IP subnets, specified by an administrator; if a
site contains more than one subnet, the subnets should be connected
by high-speed, reliable links
Flexible single master operations: Operations that can only be
performed by one specific domain controller
You must install Active Directory on all Windows 2000 Server computers that
you want to function as domain controllers — Active Directory is not installed
by default.
78 Part I Introduction to Windows 2000
Active Directory is dependent on the Domain Name System (DNS). Both
Active Directory and DNS use the same hierarchical naming conventions.
In addition, on a Windows 2000 network, client computers use DNS servers
to locate Active Directory domain controllers.
When Active Directory is implemented on a Windows 2000 network,
different types of client computers have varying degrees of functionality
with Active Directory.
When getting ready to implement Active Directory on your Windows 2000
network, there are several elements you should consider planning for, including
your domain design, naming conventions, and how your existing client computers
will fit into your overall Windows 2000 Active Directory implementation plan.
Chapter 2 Overview of Active Directory 79
STUDY GUIDE
This section contains several exam readiness questions designed to test
your knowledge of Active Directory terms and concepts and help you
prepare for the Directory Services exam.You can find the answers to these
questions at the end of this chapter.
Assessment Questions
1. Which of the following is not a feature of Active Directory?
A. It is flexible and extensible.
B. It is scalable to any size network.
C. It provides ease of administration by utilizing group policies.
D. It eliminates the need for trust relationships between domains.
2. Which of the following are classes of Active Directory objects?
(Choose all that apply.)
A. User
B. Group
C. Domain
D. Workgroup
E. Organizational Unit
3. What is the minimum number of domains that a domain tree
can contain?
A. 1
B. 2
C. 3
D. 4
4. By default, what type of trust are all Windows 2000 trust relationships
within a domain tree or forest?
A. Explicit trust
B. One-way trust
C. Transitive trust
D. Non-transitive trust
80
5. Which master role causes the domain controller that performs this
role to be the only domain controller in the forest that can add a new
domain to the forest?
A. PDC emulator
B. Schema master
C. Relative ID master
D. Infrastructure master
E. Domain naming master
6. For most large companies, in terms of ease of network administration,
what is the optimum number of Active Directory domains to use on
their Windows 2000 network?
A. 1
B. 2
C. 3
D. More than 3
7. Which of the following are true statements about organizational units
(OUs)? (Choose all that apply.)
A. They are security principal objects.
B. They are sometimes called container objects.
C. They should mimic the company’s organization chart.
D. They should be used to make network administration easier.
E. They can contain objects and other organizational units from
their own domain.
8. You want to implement Active Directory on your Windows 2000
network.Your network consists of Windows 2000 Server computers,
Windows 2000 Professional computers,Windows NT Workstation
4.0 computers, and Windows 98 computers.You want to achieve
optimum functionality of all of the client computers with Active
Directory.What should you do?
A. Install the Directory Service Client on all of the Windows 98
computers.
B. Upgrade all of the Windows 98 computers to Windows 2000
Professional.
81
C. Upgrade all of the Windows NT Workstation 4.0 computers and
Windows 98 computers to Windows 2000 Professional.
D. Nothing.Windows 2000 will automatically detect all client
computers and optimize them to function with Active Directory.
Answers to Chapter Questions
Chapter Pre-Test
1. The correct answer consists of any two of the following key Active
Directory features:
It provides fully integrated security.
It provides ease of administration by utilizing group policies.
It makes resources easier to locate.
It is scalable to any size network.
It is flexible and extensible.
2. Domains are the fundamental units that make up Active Directory.
3. A domain tree is a hierarchical grouping of one or more domains
that must have a single root domain, and may have one or more
child domains. In contrast, a forest is a group of one or more domain
trees, linked by transitive trusts, that shares a common schema and
global catalog.
4. Active Directory is dependent on DNS.
5. The primary purpose of OUs is the organization of related objects
and other organizational units to simplify administration.
Assessment Questions
1. D. Active Directory does not eliminate the need for trust relationships.
2. A, B, C, E. Of the items listed, only “Workgroup” is not a formal
class of Active Directory objects.
3. A. A domain tree is a hierarchical grouping of one or more domains
that must have a single root domain, and may have one or more
child domains.
82
4. C. By default, all Windows 2000 trusts within a domain tree or forest
are transitive trusts.
5. E. The domain naming master is the only domain controller that
can add or remove domains to/from the forest.
6. A. Using a single domain greatly simplifies the administration of
your network.
7. B, D, E. Organizational units (OUs), which are sometimes called
container objects, can contain objects and other OUs from their
own domain, and should be used primarily to make network
administration easier. OUs are not security principal objects,
and should not mimic the company’s organizational chart unless
that is how the network is administered.
8. C.To achieve optimum functionality of the client computers, you
must upgrade all of them to Windows 2000. Upgrading some of
them and/or installing the Directory Service Client will gain some
functionality, but the question specifically states that “optimum
functionality” is the required result.
83
|
Responses
|
No responses found. Be the first to respond and make money from revenue sharing program.
|
|
|