I would want to know if SynonymTokenFilterFactory in SynonymTokenFilterFactory.java uses which kind of SynonymMap. Is it FST based or is it a normal Map builder. Asking this question after having a look at https://issues.apache.org/jira/browse/LUCENE-3233
I am facing issues with large synonyms files (almost 300MB) and the same analyzer is being used in multiple indices in the same cluster. As the cluster state becomes too big because of the synonym files. Its leading to memory pressure and GCs going full.