Storing prerecorded content

Applies To: Windows Server 2008, Windows Server 2008 R2

Content storage only becomes a real concern when you are managing a large number of digital media files. In many cases, you can maintain all of the relevant digital media files in a directory on the server itself. As your content library grows, you may need to develop a separate file storage and management solution.

  • Naming. A file naming convention is one of the most useful content management techniques you can employ. For example, you can use alphanumeric codes to specify such things as digital media type, genre, artist, and sequence number. If used consistently, a well-designed file naming standard will enable you to manage virtually any number of files effectively.

  • Folders and servers. Dividing your digital media files into separate folders lets you cluster your content according to your own criteria. Keep the number of folders to a minimum to avoid confusion and redundancy.

  • Archives. Maintaining an up-to-date digital media library requires not just the addition of new content as it becomes available, but also the archival of old or obsolete content. It is usually not practical or necessary to delete inactive content, but inactive content should be removed from the active digital media library and stored in a way that will allow you to retrieve it at a later time. Use file compression technology if necessary to reduce file sizes to manageable levels.

  • Backup and security. Your content library, like any other repository of data, is vulnerable to damage or theft over the local network. Refer to your network documentation or ask your network administrator about setting up a system of user permissions and periodic backups for your content library.

  • Remote storage. The default location for storing content that you want to host on a Windows Media server is the local hard disk of the server computer. To make content available for streaming to users, a folder or file on the drive is mapped to an on-demand publishing point. This method of storing content works for many streaming media sites. However, local storage has its drawbacks if a streaming site does one or more of the following:

    • Hosts many concurrent connections. In addition to processing many concurrent streaming requests, the server must handle basic hard disk I/O operations and other operating system tasks. This can lead to high memory and CPU usage, and result in a slow response to client requests or even dropped connections.

    • Hosts many large files. The local hard disk may simply not be large enough for the number and size of source files. With one local drive handling all server tasks there may also be no provision for fault tolerance. If the drive fails, so does the server, operating system, and all the content.

    • Draws content from many sources. In many scenarios, the volume of users accessing content may be low to moderate, but administering one centralized server may not be the most efficient way to manage content.

    In a heavy streaming scenario, there are a number of drawbacks to local storage. Solutions that use SCSI devices, such as external RAID arrays, work fine to a point. However, a number of very high volume Internet sites have turned to network attached storage (NAS) and storage area network (SAN) solutions to provide the scale they need to handle the volume now and in the future. For more information, see Sourcing Content from Remote Storage.