In this chapter, we learned how to work with applications in a modern distributed server architecture, to allow for scalability and better performance. We have seen a working cluster implemented with a filesystem-based directory provider and JMS-based backend, and now have enough knowledge to explore other approaches involving Inifinispan and JGroups. We used sharding to split a Lucene index into smaller chunks, and know how to go about implementing our own custom sharding strategy, if necessary.

This brings us to the end of our little adventure with Hibernate Search! We have covered a lot of critical concepts about Hibernate, Lucene and Solr, and searches in general. We have learned how to map our data to search indexes, query and update those indexes at runtime, and arrange it all in the best architecture for a given project. We did all of this through an example application, that grew with our knowledge from simple to advanced along the way.

There's always more to learn. Hibernate Search can work with dozens of Solr components for more advanced functionality, and integrating with a new generation of "NoSQL" data stores is possible as well. However, you are now equipped with enough core knowledge to explore these horizons independently, if you wish. Until next time, thank you for reading! You can find me online at steveperkins.net, and I would love to hear from you.