One basic requirement for an identity management system is the ability to import and process identity data from an external system. This guide walks you through the main building blocks that are involved in the process of populating Microsoft Forefront™ Identity Manager (FIM) 2010 with group data from Active Directory® Domain Services (AD DS), outlines how you can verify whether your scenario works as expected, provides suggestions for managing Active Directory groups by using FIM, and lists additional sources for information.
In this section, you will find information about the scope of this document. In general, "How Do I" guides are targeted at readers who already have basic experience with the process of synchronizing objects with FIM as covered in the related Getting Started Guides.
This guide is intended for information technology (IT) professionals who already have a basic understanding of how the FIM synchronization process works and are interested in getting hands-on experience and more conceptual information about specific scenarios.
This document assumes that you have access to a running instance of FIM 2010 and that you have experience in configuring simple synchronization scenarios as outlined in the following documents:
The content in this document is scoped to function as an extension to these introductory documents.
The scenario outlined in this document has been simplified to address the requirements of a basic lab environment. The focus is to give you an understanding of the concepts and technologies discussed. This document helps you develop a solution that involves managing groups in AD DS by using FIM.
The procedures in this document require 120 to 150 minutes to complete. These time estimates assume that the testing environment is already configured and does not include the time required to set up the test environment.
If you have questions regarding the content of this document or if you have general feedback you would like to discuss, feel free to post a message to the Forefront Identity Manager 2010 forum.
Fabrikam, a fictitious company, is planning to use FIM to manage the group accounts in the corporation’s AD DS by using FIM. As part of this process, Fabrikam needs to synchronize group data to FIM. To start with the initial testing, Fabrikam has installed a basic lab environment that consists of FIM and AD DS. In this lab environment, Fabrikam is testing a scenario that consists of a security group that was manually created in AD DS.
The objective of this scenario is to synchronize the group to FIM.
To use this guide, you need three architectural components:
The following illustration outlines the required environment:
You can run all components on one computer.
The following table lists the components that are part of this scenario in this guide.
The scenario outlined in this guide consists of the following building blocks:
In this section, you will find instructions for the resources that you need to create that are outside of your FIM environment.
You need the organizational unit as a container for the sample objects.
For the scenario in this guide, you need two Active Directory user accounts:
In both cases, it is sufficient to create regular user accounts. More information about the specific requirements of both accounts is found later in this document.
For the configuration steps in this section, you need to start the FIM Synchronization Service Manager.
For the scenario in this guide, you need to create two management agents:
When you configure a management agent for AD DS, you need to specify an account that is used by the management agent in the data exchange with AD DS. You should use a regular user account. However, to import data from AD DS, the account must have the right to poll changes from the DirSync control. If you want your management agent to export data to AD DS, you need to grant the account sufficient rights on the target organizational units. For more information about this topic, see Configuring the ADMA Account.
When you import group data from AD DS, you should at least select the following attributes:
The following table lists the most important scenario specific settings you need to configure:
When you configure a FIM Service management agent, you need to specify an account that is used by the management agent in the data exchange with the FIM Service. You should use a regular user account. The account must be the same account as the one you specified during the installation of FIM. Using Windows PowerShell to Do a FIM MA Account Configuration Quick Test contains a script that you can use to determine the name of the FIMMA account name that you specified during setup and to test whether this account is still valid.
The following table lists the most important scenario specific settings you need to configure.:
The following table lists the run profiles you need to create for the scenario in this guide:
For the scenario in this guide, you only need to configure an inbound synchronization rule. The following section provides information about the configuration of the synchronization rule.
When you create the inbound synchronization rule for your Active Directory groups, you need to add a flow mapping for the domain attribute. Populating the domain attribute is a challenge because "domain" is not an attribute of a group. When this attribute is required in AD DS, the directory service has to look up the value from the configuration container.
The following illustration shows an example of a domain partition in the configuration container:
One method used to populate the domain attribute is to implement a lookup table that determines the attribute value based on the current SID of an object. The SID attribute is a good attribute for this purpose because the value of this attribute only changes when the domain membership of an object changes.
The SID attribute of an object in AD DS consists of the domain SID plus an extension called relative identifier (RID), the unique identifier of an object within the domain database. If you know what the value of the domain SID is, you can use this value in comparison with the SID value of an object to determine the value of the domain attribute as a custom expression.
FIM provides a built-in function you can use to translate a binary SID into a string representation. The name of this function is ConvertSidToString. This function returns the string representation of a SID as domain SID + RID. For an equality comparison (Eq) of a user's SID and the domain SID, you need to remove the RID part from the user's SID. Because you know the value of the domain SID, you also know the domain SID length . You can use the length of the domain SID to calculate the part of a user's SID that you need for an equality comparison.
The following example outlines how you can use the domain SID and the user's SID to calculate the domain value with a custom expression in FIM.
A lookup of Fabrikam's SID returns a value of "S-1-5-21-4220550486-1538840966-3184992408". The SID string has a length of 41.
The first step in your custom expression is to translate the object's SID into a string representation by using the ConvertSidToString method:
ConvertSidToString(objectSid)
From this string, you only need the first 41 characters from the left:
Left(ConvertSidToString(objectSid), 41)
The question is whether this string is equal to the domain SID:
Eq(Left(ConvertSidToString(objectSid), 41)
If both values match, you can flow "FABRIKAM" as the domain name into the metaverse. If the values do not match, you should flow something like "Unknown":
IIF(Eq(Left(ConvertSidToString(objectSid),41),"S-1-5-21-4220550486-1538840966-3184992408"),"FABRIKAM","Unknown")
The FIM ScriptBox provides a script that automatically calculates the required custom expression string for the domain attribute value calculation. You can find this script here: Using PowerShell To Generate The Custom Expression For The Domain Attribute Flow The script requests the required information from a target domain controller, translates the domain information into a CustomExpression and stores the result in the clipboard.
In AD DS, each group has a type and a scope. The group type is either security or distribution. Each group type can have three different scopes:
The following illustration shows the related configuration dialog for groups in AD DS:
The FIM schema defines two separate attributes to track the type and the scope information of a group. However, in AD DS, only one attribute, groupType, is used to track this information. When you configure an outbound synchronization rule for AD DS, you need to configure an outbound attribute flow mapping that merges the values for the type and the scope into one attribute value.
To calculate the required groupType value, you can use the following table:
In FIM, the type and the scope of a group are tracked in separate attributes. When you synchronize a group from AD DS to FIM, you need to calculate the FIM attribute values from the groupType attribute in AD DS.
The groupType attribute in AD DS is a bit vector, which means that the same bits are used in the security and the distribution group to define the scope of a group. When you need to determine the scope of a group, you can simply do this by applying a BitAnd operation to the groupType attribute. The following table lists the calculation results of a BitAnd, the groupType, and a value X.
Because the bits 2, 4, and 8 are used in security groups and distribution groups to define the scope, you can use the bit mask of 2 + 4 + 8 to calculate the type of a group. If a BitOr operation of the groupType and the mask value 14 returns a 14, the processed object is a distribution group.
When you import unmanaged group information from AD DS into FIM, you need to initialize the membershipLocked attribute. The best practice recommendation is to set this attribute to false.
The next attribute that, you need to initialize is the membershipAddWorkflow attribute that should be set to Owner Approval.
The following table shows the configuration of the related inbound synchronization rule:
The objective of the initialization phase is to bring your:
To synchronize group objects in your environment, you need to enable the following management policy rules:
The following table lists the run profiles that are part of the initialization phase.
The objective of this section is to test your actual configuration. To test the configuration, you:
The following table lists the properties of the sample security group:
Before you start a first synchronization cycle for a test object, you should track the expected state of your object after each run profile that you run in a test plan. Your test plan should include next to the general state of your object (created, updated, or deleted) also the attribute values that you expect. Use your test plan to verify your test plan expectations. If a step does not return the expected results, do not proceed with to the next step until you have resolved the discrepancy between your expected result and the actual result.
To verify your expectations, you can use the synchronization statistics as a first indicator. For example, if you expect new objects to be staged in a connector space, but the import statistics returns no "Adds", there is obviously something in your environment that does not work as expected.
While the synchronization statistics can give you a first indication of whether your scenario works as expected, you should use the Search Connector Space and the Metaverse Search feature of the Synchronization Service Manager to verify the expected attribute values.
To synchronize the group to FIM, follow the steps below:
To accomplish these tasks, you run the following run profiles.:
After the delta import from AD DS, the synchronization statistics report one new object:
The objective of the delta synchronization run on your Fabrikam FIMMA is to perform several operations:
To verify that your sample user has been synchronized to FIM, open the related object in the FIM Portal.
Because you have successfully synchronized a group object from AD DS to FIM, you are now ready to add more components to your synchronization logic. As a first step, you should add group members to your security group.
The membership in a group is tracked in an attribute called a member, a multi-valued reference attribute. When you synchronize reference attributes in FIM, you need to ensure that both objects, the referencing attribute as well as the referenced attribute, are available in all layers of the synchronization service. The FIM Synchronization Service preserves existing reference relationships and also enforces referential integrity. This means that the FIM Synchronization Service ensures that the references of referencing objects are pointing to valid objects. The following illustration outlines this process for the member attribute of a group.
Keeping references intact across the various data layers (connector space, metaverse, and external system) involves the transformation of the reference value into a format that is used by each layer. For example, in FIM, references are expressed as GUID values. However, in AD DS, reference values are implemented as DNs. During a synchronization run and also during an import from and an export to a data source, the synchronization engine applies the necessary transformation of the reference values. While groups can contain groups as members, also known as group nesting, a group can also contain users as members. To preserve the references that point to user objects, you need to extend your synchronization logic to the components that are required to synchronize user objects. For the required deployment instructions for synchronizing user objects in your environment, see How Do I Synchronize Users from Active Directory to FIM. You should extend your group synchronization scenario with the user synchronization logic that is outlined in this document.
After implementing the synchronization logic for users, you should add the sample user Britta Simon as a member to the security group.
After a synchronization cycle to FIM, Brita Simon becomes a member of the group in FIM:
For more information about reference attributes, see Design Concepts for Managing Reference Attributes
When you synchronize group objects from AD DS to FIM, there are two more attributes that require your attention in the FIM Service:
While you can synchronize group objects from AD DS to FIM without populating values for these attributes, they are technically required by FIM so that the group is manageable within FIM , specifically:
One option that you have is to populate both attributes based on the managedBy attribute in AD DS. However, this method may require additional updates to your AD DS because managedBy is often not populated in AD DS.
Another method is the configuration of workflows in FIM to initialize the attributes when a new group object has been imported from AD DS into FIM. Using workflows to initialize attribute values can have an impact on your environment when a large amount of objects has been imported from AD DS. You should keep this in mind when you perform bulk imports of new objects from AD DS.
The third method to initialize these values is a scripted approach based on the FIM Windows PowerShell™ cmdlets. By using the scripted method, you can retrieve a list of the affected objects, set the values that you want, and import them back into your FIM Service data store. The tradeoff of the scripted method is the required manual interaction that can be time consuming.
Synchronizing group objects from AD DS into FIM is a relatively simple task from a configuration perspective. The only bigger challenge is the population of the domain attribute because this attribute is not directly associated with an object in AD DS and needs to be calculated. In this document, you have been introduced to a method that uses an object's SID attribute value and a lookup table for the calculation of the value.
When using this method, keep in mind that a change to your domain infrastructure such as renaming existing domains or adding new domains requires an update to the calculation logic in your inbound synchronization rule. Updates to synchronization rules require full synchronization runs that can be time consuming to process in your environment.
An alternative to populating the domain attribute in a synchronization rule is the implementation of workflows to do this task. However, initializing attribute values by using workflows has an impact on the performance of your system. The performance impact of bulk imports from AD DS can be significant.
In general, when implementing inbound synchronization rules in your environment, you need to differentiate between two scenarios:
The initialization phase requires special care in your planning. In this phase, FIM is populated by a bulk import with existing objects from AD DS. This scenario can have a significant impact on both the servers in your external systems and your servers running FIM. If you have a large number of objects that you need to synchronize from AD DS into FIM, you should investigate options that can limit this process such as:
Each method has pros and cons. You need to determine the method that works best for you in your test environment.
Another aspect of handling the initialization phase is the subject of attribute flow precedence. During the initialization phase, you may need to make AD DS authoritative for attributes for which you want FIM to be authoritative after the initialization phase. This means that you need to configure the attribute flow precedence in a way that enables your Active Directory management agent to flow attribute values into FIM that, later in your deployment, will be controlled by FIM. In general, you should develop a plan that includes a solution for the initialization phase for your environment.
For the scenario in this document, the samAccountName attribute is used to join related identity objects. However, in many scenarios, this attribute is not the optimal solution because the samAccountName attribute is only required to be unique on a per domain basis. The best attribute to implement robust joins is a unique identifier called Correlation ID. There are several options to implement a Correlation ID in a production environment. You can find more information about this topic in Design Concepts for Correlating Digital Identities.
When you are in the process of planning your deployment, you should investigate an option to extend the Active Directory schema with an attribute that stores a Correlation ID. If you want FIM to eventually be authoritative for the object management in your environment, you can use the already existing GUIDs of the FIM objects as the source for your Correlation ID. In addition to implanting a method for robust joins in your environment, a Correlation ID is helpful to detect managed objects in your AD DS. If an object has a value for this attribute, you know that this object is managed by FIM.