For many use cases, performance will require little or no attention. On the JVM in particular Clara has been heavily optimized and successfully used in both cases with thousands of rules and hundreds of thousands of facts and with much smaller numbers of rules and data sets. As a rules engine based on the RETE algorithm, it broadly speaking has the performance characteristics one would expect from that algorithm. However, as with all software, there are cases where performance does become an issue, and this section of the documentation is intended to provide information on how to diagnose and remedy performance problems. Individuals topics are listed in the sidebar to the right. More topics will be added over time.