As per the documents, we need Cluster.initial_master_nodes only for the first bootstrap of the cluster.
How does a node join the cluster when it restarts? Does it only need the sed_hosts to discover others?
What if the entire cluster went down? will the nodes use the initial_master_node setting?
Since in most cases, we gradually add nodes to the cluster the initial_master_nodes setting should always be the very first node that started. In what scenarios will you add 2 or more nodes names to this setting?
No, it's only used the very first time the cluster starts.
It's ok to start your nodes one-at-a-time if you want, but it's not unusual to start up all the nodes at once to make sure that it's fault-tolerant right from the start.
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.