Export (0) Print
Expand All

FSRM and FCI: Frequently Asked Questions

The Windows Server 2008 R2 File Classification Infrastructure (FCI) automates classification processes so that you can manage your data more effectively. You can save money and reduce risk by storing and retaining files based on their business value or impact. The built-in solution for file classification provides expiration, custom tasks, and reporting. The extensible infrastructure enables you to meet additional customer classification needs by building rich end-to-end classification solutions that are built on the classification foundation of Windows Server in a consistent and supported way and within the existing Windows file serving platforms.

The File Classification Infrastructure (FCI) is controlled through the File Server Resource Manager (FSRM) interface and is installed along with the FSRM role service.

See “Creating a file expiration task” in the Working with File Management Tasks section of the File Server Resource Manager Step-by-Step Guide for Windows Server 2008 R2 (http://go.microsoft.com/fwlink/?LinkID=160157).

The file remains classified if the file is moved to another NTFS file system by using a standard mechanism such as Copy or Move. However, the properties for files in Microsoft® Office® products remain attached regardless of how the file is moved.

The properties are stored in an alternate data stream, which is a feature of NTFS. Alternate data streams move with a file if the file moves within NTFS file systems, but they do not appear in the file’s contents. The properties are also stored within file formats in Office products as custom document properties or server document properties.

The speed at which files are classified depends on the classification rules that are defined, the classification mechanisms in use, and the types of data. Approximately 40 files are classified per second when the content classifier is used to classify Office 2007 documents that are under 1 megabyte (MB). When the folder classifier is used, approximately 800–1000 files are classified per second.

File Classification Infrastructure (FCI) and SharePoint 12 store properties for documents in Microsoft Office in the same way. When a classified document is uploaded to SharePoint, the property transfers with it. If a column in a SharePoint document library is configured with the same name and an equivalent type, the property is exposed correctly in SharePoint. If the property is modified in SharePoint and the file is downloaded to a file server, FCI can read and extract the updated property.

You can use an application programming interface (API) to create new file classifications and to store properties. You can create applications or scripts that can be initiated by file management tasks in FSRM.

For more information, see IFsrmPipelineModuleImplementation Interface (http://go.microsoft.com/fwlink/?LinkId=153167).

FCI classifies files on a scheduled basis. However, you can apply real-time classification by writing an application that reacts when files are created or modified and then passes the correct parameter to the FCI APIs.

For more information, see FsrmGetFilePropertyOptions Enumeration (http://go.microsoft.com/fwlink/?LinkId=153172).

The GetFileProperty and EnumFileProperties APIs allow you to retrieve all properties that are currently known for a file, regardless of whether the current server defines all of those properties.

To alter FCI configuration on multiple servers, you write a script by using the FCI API that makes the correct modifications. You can use Group Policy to launch the completed script.

The FCI APIs that access properties exist only in the Windows Server 2008 R2 operating system. However, documents that are classified in Office 2003 and Office 2007 products are listed as Document Properties in the user interface.

For information about moving classified files, see the section “What happens to a file’s classification properties when the file is moved?” earlier in this topic.

See “Creating a file expiration task” in the Working with File Management Tasks section of the File Server Resource Manager Step-by-Step Guide for Windows Server 2008 R2 (http://go.microsoft.com/fwlink/?LinkID=160157).

There are no cmdlets for FCI. However, all FCI APIs are Component Object Models (COMs) that you can automate by calling them from a scripting language such as Windows PowerShell or JScript.

Soft links are not classified; however, the target file is classified if it is within the scope of a classification rule. Hard links are classified, and the target file stores the classification properties. If hard links cause a file to fall within the scopes of multiple classification rules, it can take multiple runs to set all the classification properties for the file.

No. FCI only classifies files locally on the server. It also classifies folders that are stored on a mounted drive.

Yes. You need to install Windows TIFF IFilter on the server that is running FCI. Then the content classifier can recognize TIFF images and extract text from those files. Other file types require other extensions.

For more information about Windows TIFF IFilter, see the Windows TIFF IFilter Installation and Operations Guide on the Microsoft Web site (http://go.microsoft.com/fwlink/?LinkId=150226).

You can add conditions against the classification properties that are defined on the server to select files for a file management task to operate on.

You can use the ClearFileProperty API to remove a classification property from a file.

For more information, see IFsrmClassificationManager::ClearFileProperty Method on the Microsoft Developer Network (http://go.microsoft.com/fwlink/?LinkId=153182).

When you restore the Cluster service that is running FSRM, an authoritative cluster restore is required. In other situations, a simple system state recovery is enough.

Errors that are encountered during the classification process are logged in the StorageReports directory. The log should contain enough information to help you resolve the situation.

Errors that are encountered during file management tasks are logged in the StorageReports directory. The log should contain enough information to help you resolve the situation. However, the log does not indicate whether a custom action completed successfully. To determine this, the custom action has to produce a separate error log.

The scope of a file expiration task should not be configured so that it could expire the target expiration directory. Such a configuration could recursively expire the files and their directory structures. (The directory structure expands each time the file expiration task runs, resulting in extremely deep directory structures.) When the operating system reaches the character limit for the path name, the expiration process terminates. The resulting directory cannot be deleted by using Windows Explorer.

FCI attempts to aggregate the property values.

  • For Yes or No properties, a Yes value takes priority over a No value.

  • For ordered list properties, the highest property value takes priority.

  • For multiple choice properties, the property sets are combined into one set.

  • For multiple string properties, a multistring value is set that contains all the unique strings of the individual property values.

  • For other property types, an error occurs.

FCI does not identify individual files within a container file. However, a non-Microsoft classification plug-in can attempt to assign an overall classification to the container file that is based on the files within the container.

File management tasks with conditions against classification properties will not match files that do not have that classification property set, unless the condition is "not exist." In that case, the task selects files that do not have the property defined.

Was this page helpful?
(1500 characters remaining)
Thank you for your feedback

Community Additions

ADD
Show:
© 2014 Microsoft