Interaction History component processes data longer than expected
The Interaction History component in a strategy takes more time to process data than expected.
This behavior can be caused by one or more of the following conditions:
- The FACT database table contains an excessive number of records. As the transactional data in this table grows, processing time slows down.
- The database that stores interaction history records is not optimized for quick retrieval of data.
Perform one of the following steps to increase the speed with which the Interaction History component processes data:
- Use the pxOutcomeTime property to archive old FACT
For example: You can archive all FACT records that are older than 31 days if your decision logic indicates that insurance policies become invalid after one month. For more information, see Applying sample scripts for archiving and purging.
- Improve the data retrieval from the data set that contains the
- Retrieve the interaction history SQL queries so that you can analyze their format and estimate the processing time. For more information, see Starting a database trace.
- Based on the retrieved queries, create a database index on only the problematic columns to quickly locate data without having to search every row in a database when that database is accessed.
Previous topic Troubleshooting interaction history Next topic pyApplication and pyApplicationVersion properties are not captured in the interaction history