An archiving architecture is typically composed of distinct functional components: an archive agent installed on the application or file server that scans data against defined criteria and moves qualifying files, an archive storage device that physically holds the fixed content, and an archive server that acts as the central management point where administrators define and configure the policies governing what gets archived, based on criteria such as file age, type, size, or last-access date, and when. In this architecture, storing the fixed data (A) is the responsibility of the archive storage device, not the archive server itself. Scanning and archiving files (D) is executed by the archive agent running close to the data source, which applies the policies that the archive server has defined, rather than performing the policy configuration itself. 'Creating policies based on file names' (C) narrows the archive server's role to a single, overly specific criterion, when in practice policies can be defined using multiple attributes beyond filename alone, making this option incomplete rather than descriptive of the server's actual function. The archive server's defining role in the architecture is the centralized configuration and management of archiving policy.

Reference topic: Replication and Data Archiving - Archiving Architecture and Components.