I need to create an Elasticsearch cluster on Ubuntu machines, but to ensure high availability I would like to have more than one node in case I need to update or upgrade the server.
Is there any material that teaches how to create this type of structure from scratch? Because I search and I don't know how to configure Elasticsearch in this type of environment.
It will never be accessible via the internet, only my VMs running on my hosting provider.
I will be working with data that may have been created months ago and retrieved for searches and reports, but more current data from the last 14 days will always be used for charts.
What is recommended in this case, using 2 or 3 nodes?