How Discover Works

Discover software running on computers in your data center is configured to see every byte of data exchanged between your web servers and the browsers on your visitors' computers. Discover passively captures the bi-directional data stream and forwards a copy of each relevant packet of that data to a server running the Discover system.

The folowing lists elaborates Unica Discover setup with three types of Discover servers.

  • Capture Server - Discover's Network Capture software reassembles TCP/IP packets into HTTP requests and responses for each exchange between visitor and web application. Optionally, uninteresting data can be dropped, and sensitive data can be deleted. The captured combinations of requests and responses (called hits) are then forwarded to another server running the Discover processing software.
  • Processing Server (also called Canister Server) - Individual hits sent from the Capture Server are grouped together into a Discover session, which is assembled hit-by-hit to include all web page interactions between a specific visitor and your website. The data contained in this session is scanned for keywords and codes that you have defined. When matches are found, a record of this event is stored for additional processing, including generation of reports and alerts. When the session is complete, the session is indexed for search and written to disk.
  • Discover Portal (also called Report Server) - This server's Web-based interface allows data analysts, business owners, IT staff, and administrative users to view the status of interactions with your website, with aggregate data reports that display what visitors are doing on your site. Discover users can search the saved visitor sessions and analyze them for common attributes, such as order completion. Individual visitors' sessions can be replayed with the actual data that was sent and received, so a high-fidelity recreation of the visitor's experience can be reviewed.