Understand sites

1
Goals

• Understand sites
• Design a site structure
• Understand operations masters
• Place domain controllers

2
(Skill 1)
Understanding Sites

• Sites
• Logical representations of physical structure
• Locations where all computers are well-connected
• Well-connected means that all computers in a
  single site are connected by high-speed,
  reliable, and cost-effective links (such as a
  LAN)
• Sites and site links have three main purposes
• Controlling replication
• Providing local logons
• Controlling certain directory-enabled
  applications (Dfs)

3
(Skill 1)
Understanding Sites (2)

• Controlling replication
• Types of replication
• Intrasite
• Occurs spontaneously based on database changes
• Allows for fast database convergence, but can be
  chaotic and bandwidth-intensive
• Intersite
• Occurs between designated bridgehead servers
  according to a schedule and interval
• Always compressed
• Path can be influenced by modifying the costs
  associated with site links

4
(Skill 1)
Understanding Sites (3)

• Site link properties
• Sites participating in the link
• Schedule upon which the link should be used
• Interval in which replication will occur over the
  link
• Cost for the link

5
(Skill 1)
Figure 10-2 an example site link
6
(Skill 2)
Designing a Site Structure

• Site design
• An important aspect of overall Active Directory
  design
• Components
• Defining sites
• Associating subnets with sites
• Defining site links
• Defining site link bridge
• Defining preferred bridgeheads

7
(Skill 2)
Designing a Site Structure (2)

• Defining sites
• In most networks, each physical location is its
  own site
• If links between locations are fast, reliable,
  and inexpensive, may be able to combine more than
  one physical location into a single site

8
(Skill 2)
Designing a Site Structure (3)

• Associating subnets with sites
• Subnets are used to allow clients to determine
  the site to which they are connected in order to
  correctly locate local servers
• Create and configure site links
• Mirror physical connectivity
• Specify site link properties cost, schedule, and
  interval

9
(Skill 2)
Designing a Site Structure (4)

• Site link schedule
• To define, analyze the physical link(s) between
  sites with a monitoring application to determine
  periods of peak activity and amount of available
  bandwidth during period
• If link(s) have a low amount of available
  bandwidth during peak periods, configure a
  schedule for the site link that excludes peak
  periods

10
(Skill 2)
Designing a Site Structure (7)

• Site link bridging
• Automatically enabled by default
• Allows direct replication to occur between sites
  that do not have a dedicated site link

11
(Skill 2)
Designing a Site Structure (8)

• Preferred bridgeheads
• Bridgeheads are the focus of intersite
  replication
• Knowledge Consistency Checker
• Creates connection objects only between
  bridgehead servers for intersite replication
• You can provide KCC with a list of one or more
  preferred bridgehead servers

12
(Skill 2)
Figure 10-8 Automatic site link bridging allows
replication from site A to site C
13
(Skill 2)
Figure 10-10 Specifying a preferred bridgehead
server
14
(Skill 3)
Understanding Operations Masters

• Flexible Single Masters of Operations (FSMOs)
• Commonly referred to as operations masters
• Two general types of functions
• Forest-wide operations master roles
• Schema master
• Domain naming master
• Domain-wide operations master roles
• Primary domain controller (PDC) emulator
• Relative identifier (RID) master
• Infrastructure master

15
(Skill 3)
Understanding Operations Masters (2)

• Flexible Single Masters of Operations (FSMOs)
• When you create the first domain in a new forest,
  all five operations master roles are assigned to
  the first domain controller in the domain
• In child domains, only the domain-wide operations
  masters roles are assigned to the first domain
  controller in the domain

16
(Skill 4)
Placing Domain Controllers

• Domain controller placement
• Goals
• Ensure that logons are possible regardless of any
  single point of failure
• Ensure that logons are performed locally whenever
  possible
• Each site should contain at least one domain
  controller
• To ensure local logons in case of any single
  point of failure, need two domain controllers

17
(Skill 4)
Placing Domain Controllers (2)

• Also analyze need for supporting servers
• Global catalog server
• DNS servers
• DHCP servers
• WINS server

18
(Skill 4)
Placing Domain Controllers (3)

• Placement of FSMO roles
• Assign two forest-wide roles to a high-uptime
  server
• Assign schema master and domain naming master
  roles to a single domain controller
• Always have at least one additional domain
  controller to act as a standby operations master
• Assign both the RID master and PDC emulator roles
  to the same domain controller
• Do not assign the infrastructure master role to a
  domain controller that is also a global catalog
  server if there is more than one domain

19
(Skill 4)
Figure 10-11 Planning forest-wide operations
master roles
20
(Skill 4)
Figure 10-12 Planning domain-wide operations
master roles