From the documentation:
Indexing and searching your data is CPU-, memory-, and I/O-intensive work which can put pressure on a node’s resources. To ensure that your master node is stable and not under pressure, it is a good idea in a bigger cluster to split the roles between dedicated master-eligible nodes and dedicated data nodes.
Our cluster currently consists of 13 nodes, where all nodes are data nodes and nodes 1-4 are master-eligible nodes. We're having performance issues so I'm thinking that maybe it's time to remove the data role from the master-eligible nodes (according to the quoted text above).
Question: If we have four master-eligible nodes that are not data nodes, what use will we have of the three master-eligible nodes that are not acting as the master? Will they add value to the cluster during this time or only be of value as fail-over if the current master goes down and a new master needs to be elected? Will they basically be idle all of the time when another node is acting as master?