In this chapter we saw how to update Lucene indexes manually, one entity object at a time or in bulk, as an alternative to letting Hibernate Search manage updates automatically. We learned about the fragmentation that accumulates from Lucene update operations, and how to approach optimization on a manual or automatic basis.

We explored various performance tuning options for Lucene, from low-latency writes to multi-threaded asynchronous updates. We now know how to configure Hibernate Search for creating Lucene indexes on either the filesystem or RAM, and why you might choose one over the other. Finally, we worked with the Luke utility to inspect and perform maintenance tasks on a Lucene index without having to go through an application's Hibernate Search code.

In the next chapter, we will look at some advanced strategies for improving the performance of your applications. This will include recapping the performance tips covered so far, before diving into server clusters and Lucene index sharding.