SharePoint 2010 Search : Setting Up the Crawler - Crawling SharePoint Sites & Crawling Users Profiles

2. Crawling Users Profiles

Enough cannot be said about the power of connecting people for business. For most organizations, their people and those people's expertise are their biggest assets. Finding people and expertise in a company can be a challenging task at the best of times, and experience and skills can go largely unexploited because people with the right knowledge cannot be found—or worse, their colleagues don't even know they exist.

SharePoint's People Search is a powerful feature to expose people in an organization and their expertise, making them findable and accessible. The people search mechanism, although a simple enough concept, requires the identification of people in the organization, their expertise, and their contact information. In order to expose this information and find the relevant people, SharePoint must first be able to crawl the information about these people.

People data in SharePoint comes from indexing user profiles. User profiles are held in SharePoint and hold the information about all the users of SharePoint as well as other potential SharePoint users that may have profile data imported from Active Directory or some other directory server service. User profile data can be entered manually, either by the administrator or by the users themselves in their personal site (MySite). Additionally, other data sources can be used to populate user profile data.

Usually the starting point for an organization is to synchronize the existing information they have in their organization's directory with SharePoint and then allow connected users to enrich that information on their MySite pages. This will allow for rich metadata and social search functionality in People Search. However, this is not strictly necessary, and data from a directory server is not required to have a rich people search experience as long as users are aware of the MySite feature and have the time and interest to keep it up to date.

User profile data is managed by the User Profile service application in the Service Applications section of Central Administration. We will only go into crawling user profiles and synchronizing them with Directory Servers, but it is important to note that a great deal of rich user information can be managed from this service application. Additionally, the User Profile service application makes it possible to share user data across multiple sites and farms. This can allow for a rich and effective people search and expose expertise in areas of the organization not previously accessible to many employees.

The protocol used to crawl data collected from the User Profile service is called SPS3. It can be seen set in the default content source for SharePoint sites as sps3://servername. If user profiles are not crawled, check if this site is set in the default content source.

2.1. Synchronizing User Profiles

To synchronize user profiles, navigate to the "Manage service applications" page in Central Administration (Figure 3).

Figure 3. The Application Management menu

Then choose the User Profile Service Application link (see Figure 4). Check if the service application is started in the right-hand column.

Figure 4. Choosing the User Profile Service Application link

The sources for user profile information are managed in the Synchronization section, where different user data sources can be defined and synchronization schedules defined (Figure 5).

Figure 5. The User Profile Service Application page

Selecting the Configure Synchronization Connection menu item will display the page where additional user profile data sources can be added. To add a new source to synchronization, select Create New Connection. The following source types can be selected:

• Active Directory

• Active Directory Logon Data

• Active Directory Resource

• Business Data Connectivity

• IBM Tivoli Directory Server

• Novell eDirectory

• Sun Java System Directory Server

1.1. User Accounts

The synchronization connection settings should use an account that has access to the data source. The account should be able to read all the user profile data on the directory server being synchronized. How this account is set up may vary depending on the source system. The source systems are defined on the Add Synchronization page and can be set by choosing the correct system (listed here) in the Type drop-down menu, shown in Figure 6. When using Business Data Connectivity Services, an account is not required as the accounts set in the Business Data Connectivity entity will be used.

• Active Directory: The user must have Replicate Directory Changes permission on the target domain. For Windows 2003 Active Directory installations, the user should be a member of the Pre-Windows 2000 Compatible Access group. For domains with a different NetBIOS name, the cn=configuration container in Active Directory should be checked to make sure the user has Replicate Directory Changes permission. As exporting property values is also likely, the user should also have Create Child Objects and Write All Properties permissions.

  Figure 6. Adding a new Directory Server source for synchronization

  

• Novel eDirectory: Novel eDirectory should be configured to allow the specified user to browse in the Entry Rights property for the directory tree where the users to import are listed. Additionally, it should give the user Read, Write, and Compare rights in All Attributes for the same directory tree.

• Sun Java System Directory Server: In the Sun Java System Directory Server, Anonymous Access to RootDSE should be enabled with Read, Write, Compare, and Search rights. If the intention is to incrementally synchronize (recommended for large directories), the account should also have Read, Compare, and Search permissions on the cn=changelog object.

• IBM Tivoli: The account specified to synchronize with IBM Tivoli should be a member of the administrative group.

2.2. Excluding Directory Tree Nodes

SharePoint allows the administrator to choose which levels of the directory and which nodes to synchronize. All other nodes will be excluded. The administrator can select the entire directory by choosing Select All or choose each organizational unit separately. Since, generally, current user data is interesting only to an organization, it is recommended that care is taken to select only organizational units that have current users in them. Exclude former employees, service users, computers, etc. See Figure 7.

Figure 7. Selecting the directory containers and users to be synchronized for user data

2.2.1. Searching from MySites

Each MySite has its own search box that is similar to all the pages in a SharePoint site. However, this MySite search box can have a different target search center. Designating the target search center for MySites is done in the User Profile services application under MySite Settings (Figure 8).

Figure 8. Setting the appropriate target search center for MySites