Fast Deletes of Retail Transaction Data in 24/7 Environments

Abstract


All retailers from Aldo to Kroger have one thing in common.  The accumulation of the sales transaction data generated by their customers.  


Since a transaction is the history of every product in a customer’s shopping basket as this data is stored in the retailer’s database the volume of transaction information can grow very quickly.  


Retention of transaction data varies between different types of retailers with high transaction volume retailers (i.e., grocery stores) maintaining no more than two months of transaction history within their OLTP environment before migrating the history to a Data Warehouse for data mining.


While slower volume/higher priced product retailers (i.e., department stores) maintain years of transaction history within their OLTP environment.


It doesn’t matter if a retailer has low or high transaction volumes, over time as the data grows it must be periodically purged to maintain performance requirements.


Many retailers purge transaction data using database partitioning to simply roll off unneeded transaction data. However, retailers with high transaction volumes or who cannot afford licensing costs for database partitioning are better suited running automated data deletion routines to periodically remove stagnant data.


This white paper addresses an approach that can be used to delete high volume, retail transaction data of retailers in 24/7 production environments.