Index Table Pattern

Find all customers that live in Coventry

> Duplicate Data with different keys

works best with static data

> Normalised Index Tables with ref to primary keys

> Partially normalised Index Tables with commonly accessed data duplicated

> Can implement composite keys when relevant

Considerations

> Index tables only for regular queries

> Overhead and Cost increased by data duplication management

> Index tables with ref to primary key = Double lookup

> Implement Eventual Consistency for Index Tables (ex. Windows Azure storage tables and transactional updates

Do not use:

> On volatile data

> With non-discriminating secondary keys (ex. Gender)