In past years, prior to deploying an application, significant time and effort was expended in gathering requirements to produce a design. The design was completed up front, before any code was written or database schema created. Furthermore, the design was comprehensive and static – it was expected to be viable for years without requiring significant modifications. When changes eventually were needed, they became a major redesign effort that often impacted multiple facets of the application and database. None of this was completed quickly, or without risk.
We are all aware of the inherent value of customer data within the databases that support our applications. This data identifies customers, accounts, transactional interactions and other patterns that are important for understanding customer behaviors and habits – all of which are useful for improving business processes and customer satisfaction. But what about the value of data about the data processing workloads? Is there value in understanding the processing workloads of our customers? What could we do with workload information if we had the ability to easily organize workload data into meaningful categories and groups, and view the dynamics of the processing patterns over time?