Elastic search Cluster auto fail over

Elastic Stack Elasticsearch
1

we have configured cluster of 2 boxes where both nodes will be master-eligible and data nodes and only one node will be playing the role of master node.
But we can still query getting them processed on any of nodes.
if 1st node goes down how the application will point to second master node , if there any automatic failure setup
Please gudie the setting

Note :- Here replication is perfectly working and we are calling 1st node in the appllcation.
and installation done on windows server 2012 R2

Thanks and Regards
Sateesh TV

2

Hey Aneesh,

Have you considered split-brain? This doc explains all about split brain. Three master eligible nodes are recommended for resiliency and to provide a necessary quorum to avoid split brain. If you have only 2 nodes and don't plan on adding a third, you would want to set your minimum_master_nodes to 2 to avoid split brain.

Hope that helps :slight_smile: