I am new to Elasticsearch and I want to set up a highly available Elasticsearch cluster on one machine to test and learn how to use Elasticsearch. I hope that when the master node fails, one of the other nodes will take over and become the master node. On my machine, I set up an Elasticsearch cluster consisting of 3 nodes, a master node and two data nodes, but when I stopped the master node, none of the other nodes took over and became the master node, here is my config for the three nodes (used version of elasticsearch is 7.9.2):
In addition to what David said, I suggest you read the notes in the manual on setting up a HA cluster with three nodes. That whole section of the manual is probably of interest to you.
You mean by keeping all the default values for node.data and node.master. I don't specify the master nodes and the data nodes even though I have n nodes and I let elasticsearch do it for me, is that right? I’m sorry if I ask a very stupid questions but I have no knowledge in elasticsearch, and soon we will install an elasticsearch cluster for one of our customers, and I need to quickly become familiar with elasticsearch.
You should really attend our trainings then to learn how to do the right things.
And/or buy a support for your customer so they will get a license and support.
Or deploy on cloud.elastic.co.
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.