No, there is not. Your better bet is to have a single IP that points to a load balancing IP, or to deploy to all of your master eligible nodes, and set master_only: true in the client block. It will then only execute on the elected master node, and just exit without doing anything on the other nodes.
Thanks, pointing at a single IP is the strategy I settled on. I'm deploying curator in a docker container, and the de facto way of providing config to a container is via env vars (a la the official Elasticsearch containers)
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.