When implementing an aggregate data table, which best practice minimizes disruption to other users?

Creating a new aggregated table while preserving the original table through a view is a best practice that minimizes disruption to other users. This approach allows the new table to be built and optimized without affecting the current system's operations or the users relying on the original table. By using a view, users can continue to access the original data seamlessly, ensuring that there is no interruption in service during the transition.

This method is particularly beneficial because it provides a straightforward way to test and validate the new aggregate table while maintaining backward compatibility. Once the new table is verified and ready for production use, it can simply replace the view or be updated in a controlled manner, further reducing risk to ongoing user queries and processes.

In contrast, using a temporary naming convention for the modified table could lead to confusion among users who are accustomed to the existing naming structure. Implementing a versioning system for all aggregates, while useful for tracking changes, may not directly address immediate disruption concerns. Cascading changes across all dependent tables can introduce complexity and risk, as it alters multiple components at once and may lead to unforeseen issues for users depending on those tables.