Planning Your File Connector Collections and Sources

The content of the Coveo unified index is organized in collections and each collection contains one or more sources (see Understanding Coveo .NET Components Hierarchy). Before starting to deploy the File connector, you should determine how to organize collections and sources for the content of your file shares.

Consider the following facts:

• End-users can see collections names in search interface elements, while sources are generally only visible by Coveo administrator in the Administration Tool.

  Note: The source names can appear in the search interface for example when a custom facet presents @syssource elements.

• You can configure a search interface to include a Collection facet (see Managing Built-in Facets and Related Results Appearing in a .NET Search Interface) or collection check boxes below the search box (see Activating Search Interface Options With the .NET Interface Editor) so that end-users can refine search results based on collections.

• Each search interface has a specific scope that is defined by including one or more collections in which to search (see Configuring the Scope of a .NET Search Interface).

• When you create a collection, you can set permissions on the collection by specifying users or groups allowed to search the content of the collection.

• Similarly, when you create a source, you can set permissions on the source by specifying users or groups allowed to search the content of the source.

Consider the following recommendations when planning collections:

• Separate your file share content in collections that are meaningful to end-users and that are useful to refine results.

  Example: When you have network file servers for different locations in your organization, create a collection for each file server:

  • New York file share

  • San Francisco file share

  • Houston file share

• When creating a collection, choose a name that is clear and meaningful to end-users.

• Consider creating separate collections for separate audiences when you define specific search interfaces for specific audiences.

Consider the following recommendations when planning sources:

• Create separate sources when you need different impersonators to fully crawl different file shares or file share sections.

• When you choose to index mail archive (.pst) files, create a source to exclusively crawl mail archive files and exclude mail archive files from the source that crawls all other file types within the same file share.

• Consider creating separate sources when you want to set different permissions to different sections of a file share.

• Avoid grouping local and remote servers on the same source to prevent delaying source refresh on all servers when one server stops responding.

Note: Use the Desktop connector when you want to index files (including mail archives) located on hard drives in end-user desktop and laptop computers (see Desktop Connector).

What's Next?

Review the deployment process (see File Connector Deployment Overview).