Thank you. From what I can see, the threads are not doing anything. They are just waiting for work. These threads are necessary for the operation of Hibernate Search; they will be there for the whole time your application is up. Most of the time they won't do anything and can be safely ignored. Why do you think these threads are a problem? By the way, the second thread dump has org.hibernate.search.backend.impl.lucene.SyncWorkProcessor in the stack trace, which indicates you are also running a Lucene backend *on top of* the Elasticsearch backend. This typically can only happen if you have two instances of Hibernate Search running, which should only happen if you have two instances of your application running with different configuration. Is that on purpose? |