Temporary Configuration Changes

The previous steps mirror the generalized flow of data through the solution, which is outlined below.

  • All items below apply to completed sessions. Items also marked (active) apply to active sessions, too.

General Data Flow of Session Data through Unica Discover

  • Active session begins.
    • (active) Session is available for replay in Portal.
  • Session is completed.
  • Session is moved from Short Term Canister to Long Term Canister.
  • Session is indexed for search in LTC.
    • (completed) Session becomes available for search by using completed search templates in the Portal.
  • Session data is aggregated for reporting purposes.
    • (completed) Session data populates Portal reports.
  • Session data is trimmed from the Long Term Canister and, optionally, archived.

For testing purposes, you may want to shorten the intervals for some of the steps to hasten the testing process. These intervals can be configured by modifying the following configuration options.

Note: Do not make these changes on a production server. These changes should only be applied in a testing environment and should be reverted when the testing is complete.
  • Session is indexed for search in LTC: For more information about configuring indexing, see "Configuring Indexing" in the Unica Discover Configuration Manual.
    1. The control program for indexing (indexprogram) checks for sessions that need indexing that is based on the Sleep Time When No Work setting.
    2. When sessions are detected that need indexing, the indexing process begins. Indexes are not committed to disk, which enables searching until the Direct Pull Timeout setting expires or the index size exceeds the Maximum Index Size threshold.
  • Session data is aggregated for reporting purposes: By default, the Data Collector polls the Long Term Canister for data to aggregate every five minutes.
    1. If necessary, you can restart the Discover Data Collector Service through the Windows™ Services Control Panel to immediately perform a data collection. However, you must be certain that the data to be collected is already generated for this step to work.