Exchange Server 2010 : Setting Up Message Routing (part 1) - Routing Messages & Using Active Directory Sites and Site Costs for Routing

1. Routing Messages

Hub Transport and Edge Transport servers route messages received from internal and external senders to their ultimate destinations. The message is first categorized and is next processed by the store driver, delivery agents, or the foreign gateway connection handler. The component that is used depends on the ultimate destination. This section describes the stages of message categorization and the various messaging components that implement message routing.

1.1. Categorizing Messages

The categorizer is a message routing component of the Exchange Server 2010 transport service that processes incoming messages and determines what to do with these messages based on information about the intended recipients. After a message is received by an Exchange Server 2010 Hub Transport or Edge Transport server, it undergoes preliminary processing and is delivered to the submission queue. Messages then move from the submission queue through the categorizer. This process can be divided into several phases as follows:

• Initial agent processing Agents such as the Microsoft Forefront Protection for Exchange Server antivirus agent and the journaling agent process messages on Hub Transport servers.
  

• Recipient resolution The recipient email address is resolved. This determines whether the recipient has an internal mailbox or an external email address.

• Routing The categorizer’s routing component determines the message’s ultimate destination and the route to that destination. It then selects the next segment (or hop) for message relay and resolves the next hop information to a list of physical servers and IP addresses.

• Content conversion This transforms email messages into a format that is readable by the recipient and that is specific to the recipient’s email client. Conversion occurs before a message is relayed to its next hop.

• Routed message agent processing After the routing decisions for a message are made, the transport rules agent and the journaling agent are applied on a Hub Transport server. Note that the journaling agent is applied both at the initial agent processing stage and when the message has been routed. This enables any changes that are made to the message by the transport rules agent to be processed by the journaling agent.

• Message assembly and DSN generation The categorized message is assembled and moved to a delivery queue. A DSN message may also be generated during this phase.

1.2. Messaging Components

When they have been categorized, messages are processed by the store driver, delivery agents, or the foreign gateway connection handler. The processing component that is used depends on the ultimate message destination. A delivery queue is dynamically generated for each hop, and messages are queued in delivery queues after a routing decision is made. If a route to a recipient cannot be found, the messages are queued to the unreachable queue.

Exchange Server 2010 accesses configuration information stored in Active Directory to make routing decisions on a Hub Transport server. On an Edge Transport server, configuration information is stored in and accessed from AD LDS on the local server. Windows Server 2008 (or 2008 R2) and Exchange Server 2010 services create mappings of the configuration data and cache these mappings in routing tables that Exchange Server 2010 references when making routing decisions. The mappings cache is updated whenever the routing topology changes. Note that Edge Transport servers cannot cache information about Active Directory topology.

The following configuration and service components are used in message routing:

• Active Directory sites These represent routing boundaries for Hub Transport servers. A Hub Transport server delivers email messages directly to Mailbox servers, distribution group expansion servers, and source servers for connectors in its local Active Directory site. It can also route messages to Edge Transport servers that are subscribed to that site. However, when routing email messages to remote Active Directory sites, a Hub Transport server must relay these messages to a Hub Transport server in the remote site.

• Active Directory IP site links IP site links define logical paths between Active Directory sites. Exchange Server 2010 uses IP site links to determine the least-cost routing path to remote Active Directory sites.

• Send connectors These are used to send messages to other SMTP hosts. If your Exchange organization routes messages to more than one email domain, you may decide to create Send connectors that are dedicated to each address space.
  

• Delivery agents These are used to route messages to foreign systems that do not use the SMTP protocol.

• Foreign connectors These use drop directories to send messages to foreign systems that do not use SMTP protocol for message transfer. Exchange uses the configuration of Foreign connectors when making routing decisions.
  

• Routing groups All computers running Exchange Server 2010 deployed in an organization belong to a single, global routing group. This is to implement compatibility with Exchange Server 2003.

• Routing group connectors These define logical paths between Exchange routing groups. They are used when Exchange Server 2010 is deployed in an existing Exchange 2003 organization.
  

• Microsoft Exchange Transport service This service is the SMTP provider for Exchange Server 2010. A series of SMTP Receive agents are triggered by various SMTP events, and the Microsoft Exchange Transport service enables these agents to process messages as they pass through SMTP transport and to perform anti-spam, antivirus, and other tasks before messages are submitted to the categorizer.

• Microsoft Exchange Active Directory Topology service This service locates the domain controllers and global catalog servers that Exchange Server 2010 uses to retrieve configuration and recipient data from Active Directory.

• Routing tables These hold the information that the routing component uses to make routing decisions. The routing table is composed of a map of topology components and their relationship to one another.

• SMTP The SMTP protocol is used for communication when messages are relayed between SMTP servers. An SMTP server can be a Hub Transport server, Edge Transport server, or smart host. A Hub Transport server uses remote procedure call to deliver messages directly to Mailbox servers that have the same Active Directory site membership as the Hub Transport server.

• DNS Exchange Server 2010 uses the enhanced DNS client component of the Microsoft Exchange Transport service to resolve the next hop selection to a list of target server names. The standard DNS client is used to resolve that list of server names to IP addresses. Enhanced DNS also provides round-robin load-balancing functionality for Exchange 2010 Transport servers.

2. Using Active Directory Sites and Site Costs for Routing

An Active Directory site is based on the physical aspects of the network and defines which subnets in the network are connected. The aim of site design is to optimize Active Directory replication traffic. The Active Directory site represents a routing boundary for Exchange Server 2010, and Hub Transport servers make routing decisions based on Active Directory site topology.

2.1. Site Membership

By default, an Active Directory forest contains a single Active Directory site named Default-First-Site-Name. If no other Active Directory sites are created, all domain member computers in the forest are members of Default-First-Site-Name, and you do not need to configure a subnet-to-site association. If you create additional Active Directory sites, you must specify the subnets that are assigned to each site. Table 1 shows a typical site-to-subnet association.

A Domain or Enterprise administrator assigns Active Directory site membership to domain controllers and global catalog servers. Other member computers in the domain, such as Exchange servers, are assigned Active Directory site membership automatically when they are configured to use an IP address in an IP subnet that is associated with an Active Directory site. Computers within the same Active Directory site are presumed to have good network connectivity. A member server is always in only one Active Directory site.

A site-aware application, such as Exchange Server 2010, can determine the Active Directory site membership of the computer on which it is installed and of other computers in the forest and then use that information to control communication flow. When a site-aware application needs to access another server, such as a domain controller or global catalog server, it first attempts to access servers that have the same Active Directory site membership as the computer on which it runs. An Exchange Server 2010 server uses Active Directory topology for message routing and to communicate with the services that are running on computers with other Exchange Server 2010 server roles installed. The Active Directory site acts as both a routing boundary and a service discovery boundary.

The process of determining the site membership of a computer in a domain uses a series of DNS queries to compare the local IP address to defined subnets and thus determine the appropriate site membership association. To reduce the overhead associated with DNS queries, Exchange Server 2010 adds the msExchServerSite attribute to the Active Directory schema. This attribute is a property of each Exchange server object, and the value of this attribute is the distinguished name of the Active Directory site of the Exchange server. Because site membership affinity is stored as an attribute of the server object, the current topology can be read directly from the Active Directory. This also enables a site membership association for a non-domain computer, such as a subscribed Edge Transport server.

2.2. IP Site Links and Site-Link Costs

Site links are logical paths between Active Directory sites. A site-link object represents a set of sites that can communicate at a uniform cost through a specified intersite transport. Site links do not correspond to the actual paths that network packets follow on the physical network, but the cost that an administrator assigns to a site link typically relates to the reliability, speed, and available bandwidth of the underlying network. For example, an administrator would assign a lower cost to a network connection with a speed of 100 megabits per second (Mbps) than to a network connection with a speed of 10 Mbps.

By default, all site links are transitive. This means that if Site 01 links to Site 02 and Site 02 links to Site 03, then Site 01 links to Site 03. The transitive link between Site 01 and Site 03 is known as a site-link bridge.

You can configure a site link to use either IP or SMTP as the communication transport protocol. An SMTP site link is designed to provide a store and forward mechanism for replication of a limited number of data types between Active Directory sites that do not have a reliable network link. All types of data can be replicated across an IP site link, and Exchange Server 2010 uses IP site links to determine its routing topology. The routing component of Exchange Server 2010 takes into account the cost assigned to an IP site link when calculating a routing table. IP site-link costs are used to calculate the least-cost routing path to the ultimate destination for a message.

Every Active Directory site must be associated with at least one IP site link and a single IP site link named DEFAULTIPSITELINK is implemented by default. When you create an Active Directory site, you associate that site to an IP site link and you can either create additional IP site links to implement the desired topology or associate every Active Directory site with the DEFAULTIPSITELINK site link. Each Active Directory site that is part of an IP site link can communicate directly with every other site in that link at a uniform cost.

Figure 1 shows a full mesh topology that uses only the single default IP site link DEFAULTIPSITELINK. Each site communicates directly with every other site by using the same cost metric. Although more than one communication path is configured, only a single IP site link is defined.

A hub-and-spoke topolgy requires additional site links. For example, in Figure 2, four sites are connected in this topology. The central site, Site A, can communicate directly with each of the spoke sites, and the spoke sites can communicate with each other through the central site.

Exchange uses site links when determining the least-cost path but will always attempt to deliver messages directly to the destination Hub Transport server. For example, if a user in Site B in the topology shown in Figure 7-18 sends a message to a user in Site C, the Hub Transport server in Site B will connect directly to the Hub Transport server in Site C. If you want to force messages to go through Site A, you need to enable that site as a hub site.

The default cost for a site link is 100. A valid site-link cost can be any number from 1 through 99,999. If you specify more than one path, the link with the lowest cost assignment is always preferred. You can assign an Exchange-specific cost to an IP site link. If an Exchange cost is assigned to an IP site link, it is used by Exchange Server 2010. Otherwise, the Active Directory cost is used.

In most cases, existing IP site-link costs, based on network speeds, work well for Exchange Server 2010 message routing. However, if costs and traffic flow patterns are not optimal for Exchange traffic, you can assign Exchange costs by using an EMS command based on the Set-AdSiteLink cmdlet. For example, the following command sets an Exchange cost of 1,000 on the IP site link LINK0304:

Set-AdSiteLink -Identity LINK0304 -ExchangeCost 1000

In Figure 3, messages from Site 01 to Site 04 would normally be routed through Site 03 based on Active Directory IP site-link costs. However, because the Exchange cost of the IP site link between Site 03 and Site 04 has been set at 1,000, Exchange will route messaging traffic from Site 01 to Site 04 through Site 02.

Adjusting IP site-link costs can be useful when the message routing topology needs to diverge from the Active Directory replication topology. You can use Exchange costs to force all messages to pass through a hub site. You can also use Exchange costs to control situations where messages are queued because communication to an Active Directory site fails or if a network connection between sites is a low-bandwidth connection used only for Active Directory replication.

2.4. Configuring Maximum Message Size on Site Links

By default, Exchange Server 2010 does not limit the size of messages that are relayed between Hub Transport servers in different Active Directory sites. If it becomes necessary to specify size limits because of traffic considerations, you can use an EMS command based on the Set-AdSiteLink EMS cmdlet to configure a maximum message size on an Active Directory IP site link. For example, the following command sets the maximum message size on the Active Directory IP site link LINKAB to 1 gigabyte (GB):

Set-AdSiteLink -Identity LINKAB -MaxMessageSize 1GB

Exchange routing generates an NDR for any message that has a size larger than the maximum message size limit configured on any Active Directory site link in the least-cost routing path. You can use this facility to restrict the size of messages sent to remote Active Directory sites with low-bandwidth connections.

Sometimes you want to ensure that all message delivery is relayed through a particular Active Directory site, such as to comply with your Exchange organization’s internal policies. You can use an EMS command based on the Set-AdSite cmdlet to designate an Active Directory site as a hub site. If a hub site exists along the least-cost routing path for message delivery, the messages queue is processed by the Hub Transport servers in the hub site before messages are relayed to their ultimate destination. If several hub sites exist along the least-cost routing path, messages stop at each hub site along that path.

The following command designates the Active Directory site MyADSite as a hub site:

Set-AdSite -Identity MyADSite -HubSiteEnabled $true


2.6. Exchange 2010 Routing Tables

When the Microsoft Exchange Transport service starts, it calculates a set of routing tables based on a snapshot of information retrieved from Active Directory (or on an Edge Transport server from AD LDS). Routing tables determine how messages are routed to recipients. When configuration changes are made, the routing tables are rebuilt, and the new routing tables are used to route incoming messages.

Exchange Server 2010 retrieves the following configuration data from Active Directory and makes it available to the routing component on Hub Transport servers:

• Active Directory sites

• Active Directory IP site links

• Exchange servers and their relationship to Active Directory sites

• SMTP connectors

• Non-SMTP connectors (these include delivery agent connectors, Foreign connectors, and non-SMTP connectors hosted by Exchange Server 2003)

• Routing groups

• Routing group connectors

• Mailbox stores

• Public folder stores

• Public folder hierarchies

Based on this data, the routing component of the Microsoft Exchange Transport service populates routing tables. The routing table correlates the data and maps the topology. This topology map contains the following elements:

• Linked connectors map This map identifies the Receive connectors on the local server that are linked to the Send connector.

• Server map This contains all Exchange Server 2010 and Exchange Server 2007 Hub transport, Edge Transport, and Mailbox servers, in addition to any Exchange Server 2003 servers in the organization. The map includes the total cost to reach any specific server.

• Legacy server map This contains all Exchange Server 2007 Hub Transport, Edge Transport, and Mailbox servers, in addition to any Exchange Server 2003 servers in the organization. The map includes the total cost to reach any specific server.

• Message Database (MDB) map This contains all MDBs in the organization and correlates the distinguished name of each MDB to routing data that includes the total cost to reach the server that hosts a specific MDB.

• Active Directory site map This contains all Active Directory sites and a structure that holds the least-cost routing path from the local site to every other site. The map includes any hub sites along the least-cost routing path. Each routing path hop also identifies all Hub Transport servers on that site that will be used by the Enhanced DNS component. DNS is discussed later in this lesson.

• Routing groups map This contains the total cost and first hop routing group connector for the least-cost routing path from the Exchange 2010 routing group to each legacy routing group.

• Send connectors map This identifies the Send connectors configured in the organization and the source servers for each connector.

The information in the routing tables is logged to routing logs. These logs are located by default in the C:\Program Files\Microsoft\Exchange Server\V14\TransportRoles\Logs\Routing folder. A new log is generated every time the routing tables are recalculated. If a Hub Transport server is unable to contact Active Directory, routing decisions are based on the currently cached data, even though that data may not be up to date.