unfortunately the resource files (e.g. synonyms) need to be present on every data node on the cluster in the same location (e.g. the config directory). There is an open issue to load data from an auxiliary index but it is not currently worked on. Definitely on the wish-list though. In the meantime you need to rely on scripting etc... to deply the files correctly in your cluster. The files get loaded once when the index is created or re-opened (after closing), so if you update all files on a closed index and reopen it, all should be good. As an alternative you can use the Reload Search Analyzers API to reload synonyms used in search analyzers, there you need to make sure to call the API endpoint only when all files have been replaces/updated.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.