I have 3 nodes, 2 with same config and one with different config. I have installed same version of elastic 6.2.3 in all of them and have started the service again after giving same cluster names to all three and default settings. All three services are running but the nodes are behaving as single node cluster.
I tried to set zen.multicast to false but it throws error that some required plugin is not installed.
This sounds like a split-brain situation. If you have three master-eligible nodes then discovery.zen.minimum_master_nodes must be set to at least 2 to avoid this. I think you also might be missing the discovery.zen.ping.unicast.hosts setting.
Multicast discovery was moved to a plugin in 2.0 and removed in 5.0. In 6.2.3 there is no such setting.
So on all the 3 nodes, I will set minimum_master_nodes to 2. And since all my nodes are master, will unicast.hosts will list all 3 hosts on every server?
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.