Setting the retention period for transaction data - Transaction Finder

Transaction Finder Plugin Guide

Applies to
Transaction Finder
Last updated
Content type
Guides > Plugin and extension guides
Transaction Finder

To automatically delete old transaction data from your Elasticsearch database, you must set a retention period.

What you should know

  • When the Elasticsearch database reaches its flood stage watermark, the Elasticsearch indices stop accepting new data. Incoming data is queued in the SQL database until the watermark level changes. This will unblock the indices and the data recovery process moves the queued transaction to the Elasticsearch database. For more information, see Setting the trigger values for Elasticsearch low disk space behavior.
  • The retention period defines how long to keep data in the Elasticsearch database.
  • The cleanup interval is user-configurable. The cleanup process of data retention occurs every day at specified time.
  • The data for the previous month is kept until the current month is over.
  • When configuring the retention period, video storage capacity should be taken into consideration. In order to correlate transactions with video footage, the corresponding video footage must still exist on the system. Transaction data retention settings do not affect video retention settings.


  1. From the Config Tool home page, open the Plugins task.
  2. Select the Transaction Finder plugin role from the entity browser, and click the Elasticsearch tab.
  3. In the Retention period field, enter the number of months that you want to retain data for.
    NOTE: The default is 3 months. The minimum is 1 month. The maximum is 24 months.
  4. In the Cleanup time field, enter the time of day to delete old data.
  5. Click Apply.


In the following example, transaction data is kept for 1 month and cleanup occurs at 2:00 A.M.

This means that during the month of February, the data for the current and previous month are kept. On March 1, all data from the previous month of February is kept, and the data from January is deleted.