Where do I write this? I didnt find something like this in elasticsearch.yml
Also there is a brief explanation in the same guide to what is a "node":
Any time that you start an instance of Elasticsearch, you are starting a node
I didnt understand from this exactly what a node is. From my understanding cluster is a bunch of nodes and a node is a bunch if indices and the logs/data flows to the indices. Is this true?
For small clusters not under heavy load there may not be any need for dedicated master nodes. When starting out with a small cluster, it is generally sufficient to set up 3 nodes where all nodes are master eligible and also hold data, which is the default configuration. Having 3 nodes in the cluster will give you high availability as the two remaining nodes will be able to make a majority decision to elect a new master if the current master was to disappear (through crash, network partition or excessive GC).
The cluster as a whole holds indices and data. If nodes crash, the cluster will redistribute the data automatically to try to ensure all primary and replica shards are assigned.
Thanks!
I will look there.
I believe there will be about 12 machines (4 of them will be heavy duty) emitting logs through filebeat/logstash to the elasticsearch
Set up 3 dedicated master nodes. These do not do a lot of work, so do not need to be very powerful, but are essential for cluster stability. Requests should be routed directly to the data nodes.
Install Elasticsearch on each of the hosts that are to make up the cluster. Locate the elasticsearch.yml file (location depends on how you install Elasticsearch) and update the settings I linked to. As you are connecting nodes distributed across multiple hosts, make sure that you pay attention to the bootstrap checks.
First of all thank you a lot!
I am getting a much better understanding of the subject thanks to you!
Just to make sure, basically when the manual says:
On one ES there can be one node, but when there are extra ES each one of them have a node and we can connect them all together
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.