When using Solr (or several other information retrieval tools), one of the most recurring open questions is, "What are the best practices to adopt for the analysis of the entities we are working with?"

In the previous chapters of this book we had a chance to start playing with Solr, adopting a multicore approach from the beginning, simply thinking about a core as an index for every single entity. This way we could define more and more cores to explore some different index configurations for the same entity (for example, painting). But, when we played with the cities core to augment our data on the paintings core, we were already using two cores at the same time, each of them defining an index for a different entity. The point is, we are always able to index our data on different cores and manage them as independent services.

There are times, however, when it could be more useful to manage data in a single index, avoiding too much relational thinking between different cores. In the next sections, we will explore how to handle multiple entities on different indexes, or with a single index.