We should try to provide metrics, tracing, etc. in Hibernate Search, in particular so that automated tools (e.g. on OpenShift) can take advantage of that data to infer configuration changes that could potentially improve performance.