Configuring a Shared SMTP Namespace

When GAL Synchronization creates contacts from mail recipients in a source forest, it uses SMTP addresses to create a TargetAddress property for each contact. Therefore, when users in a forest send mail to a contact, the mail is delivered to the contact’s TargetAddress property, even if the user manually entered the primary reply address. To determine which TargetAddress GAL Synchronization should assign to a contact, it compares the recipient’s ProxyAddresses property to the SMTP address for which the Exchange organization is responsible. Each organization must have a unique SMTP domain namespace so that contacts receive a unique TargetAddress. If your forests do not have unique namespaces, you can add a unique SMTP address to the appropriate recipient policies for each Exchange organization that contains users to be replicated across forests. After you do this, messages sent to a contact are routed directly to the source forest, where the target address resolves to the actual mailbox and the message is delivered.

You can also route contacts on a forest-by-forest basis. When setting up management agents for GAL synchronization, you can select whether mail sent to contacts that have been imported into a forest should route back through the source forest. If you have a connector to a foreign messaging system, by default, mail that is intended for a contact is routed to the source forest (the forest that manages the connector); however, the forest administrator can change this routing configuration.

Note If Exchange 5.5 is running in the forest, ADC replicates the second proxy address to the Exchange 5.5 directory, provided that two-way connection agreements are set up.

As an example of SMTP routing in a multiple forest environment, consider two forests that each have a default recipient policy with an SMTP proxy address of Configuring a Shared SMTP Namespace

When GAL Synchronization creates contacts from mail recipients in a source forest, it uses SMTP addresses to create a TargetAddress property for each contact. Therefore, when users in a forest send mail to a contact, the mail is delivered to the contact’s TargetAddress property, even if the user manually entered the primary reply address. To determine which TargetAddress GAL Synchronization should assign to a contact, it compares the recipient’s ProxyAddresses property to the SMTP address for which the Exchange organization is responsible. Each organization must have a unique SMTP domain namespace so that contacts receive a unique TargetAddress. If your forests do not have unique namespaces, you can add a unique SMTP address to the appropriate recipient policies for each Exchange organization that contains users to be replicated across forests. After you do this, messages sent to a contact are routed directly to the source forest, where the target address resolves to the actual mailbox and the message is delivered.

You can also route contacts on a forest-by-forest basis. When setting up management agents for GAL synchronization, you can select whether mail sent to contacts that have been imported into a forest should route back through the source forest. If you have a connector to a foreign messaging system, by default, mail that is intended for a contact is routed to the source forest (the forest that manages the connector); however, the forest administrator can change this routing configuration.

Note If Exchange 5.5 is running in the forest, ADC replicates the second proxy address to the Exchange 5.5 directory, provided that two-way connection agreements are set up.

As an example of SMTP routing in a multiple forest environment, consider two forests that each have a default recipient policy with an SMTP proxy address of pointbridge.com. To set up unique namespaces, you would do the following in each Exchange organization:

• In Organization 1, add an SMTP proxy address of Org1.pointbridge.com.com to the default recipient policy.
• In Organization 2, add an SMTP proxy address of Org2.pointbridge.com.com to the default recipient policy.

In both cases, when adding the proxy address, you would select the This organization is responsible for all mail delivery to this address check box. Also, you would leave the pointbridge.com proxy as the primary address so that, when a user sends mail, their reply address is user@pointbridge.com (rather than user@Org1.pointbridge.com or user@Org2.pointbridge.com).

Another example illustrates mail flow in a hub-and-spoke topology. In this example, multiple Exchange organizations are present, but all users can be addressed in a single domain space (for example, @example.com). In this case, all external mail addressed to @example.com flows into a central hub organization called OrgA. OrgA is configured with secondary SMTP proxy addresses that represent each spoke organization. One of these addresses is @OrgB.example.com. When mail addressed to UserB@example.com arrives at OrgA, the mail resolves to the contact, and the mail is redirected to OrgB. When the message leaves OrgA, the To line is changed to the TargetAddress propertyto allow for routing, but the Reply To address remains UserB@example.com.

For the following reasons, moving recipients from one organization to another does not prevent users from replying to old e-mail messages:

• The message retains the legacyExchangeDN property so that recipients can reply to the mail.
• GAL Synchronization creates a secondary X.500 proxy address for the user who was moved so that old messages can be properly routed to the user’s new mailbox based on the legacyExchangeDN property.

For example, UserA sends mail to UserB, who is in the same organization. Later, UserA is moved to a different organization. The mail originally sent by UserA still specifies UserA’s legacyExchangeDN property. GAL Synchronization creates a contact for UserA in the old organization and assigns an X.500 address with the old legacyExchangeDN property. This allows UserB to reply to the old mail, which, in turn, is properly routed to the TargetAddress property for UserA. If a mailbox is moved many times, the list of secondary proxy addresses can potentially grow large.

. To set up unique namespaces, you would do the following in each Exchange organization:

• In Organization 1, add an SMTP proxy address of Org1.pointbridge.com.com to the default recipient policy.
• In Organization 2, add an SMTP proxy address of Org2.pointbridge.com.com to the default recipient policy.

In both cases, when adding the proxy address, you would select the This organization is responsible for all mail delivery to this address check box. Also, you would leave the pointbridge.com proxy as the primary address so that, when a user sends mail, their reply address is user@pointbridge.com (rather than user@Org1.pointbridge.com or user@Org2.pointbridge.com).

Another example illustrates mail flow in a hub-and-spoke topology. In this example, multiple Exchange organizations are present, but all users can be addressed in a single domain space (for example, @example.com). In this case, all external mail addressed to @example.com flows into a central hub organization called OrgA. OrgA is configured with secondary SMTP proxy addresses that represent each spoke organization. One of these addresses is @OrgB.example.com. When mail addressed to UserB@example.com arrives at OrgA, the mail resolves to the contact, and the mail is redirected to OrgB. When the message leaves OrgA, the To line is changed to the TargetAddress propertyto allow for routing, but the Reply To address remains UserB@example.com.

For the following reasons, moving recipients from one organization to another does not prevent users from replying to old e-mail messages:

• The message retains the legacyExchangeDN property so that recipients can reply to the mail.
• GAL Synchronization creates a secondary X.500 proxy address for the user who was moved so that old messages can be properly routed to the user’s new mailbox based on the legacyExchangeDN property.

For example, UserA sends mail to UserB, who is in the same organization. Later, UserA is moved to a different organization. The mail originally sent by UserA still specifies UserA’s legacyExchangeDN property. GAL Synchronization creates a contact for UserA in the old organization and assigns an X.500 address with the old legacyExchangeDN property. This allows UserB to reply to the old mail, which, in turn, is properly routed to the TargetAddress property for UserA. If a mailbox is moved many times, the list of secondary proxy addresses can potentially grow large.