You can take advantage of Elasticsearch clusters and auto discovery features to improve data availability and resilience. If you enabled failover on the plugin role, you must set up an Elasticsearch cluster with at least one other database node.
Many nodes, one cluster
When you use the plugin installation package to deploy Elasticsearch database nodes on multiple workstations, the name of the default cluster is the same everywhere. When nodes share the same cluster name, then when one node receives data it informs the other nodes. So, all nodes inside the cluster have the same data.
When a single node fails, data remains available
If one node gets full, fails, or otherwise becomes unreachable, the plugin continues to have access to the transaction data from the other nodes, even when the plugin role is configured to connect to the local Elasticsearch database.
When the plugin role fails over, data remains available
If the plugin role fails over to a secondary server, the plugin continues to have access to the latest transaction data from the other nodes in the cluster.
How to deploy a cluster
For detailed instructions on configuring multiple, distributed Elasticsearch nodes in a cluster, refer to the Elastic documentation for your version of Elasticsearch. The documentation for Elasticsearch 7.10 is available at https://www.elastic.co/guide/en/elasticsearch/reference/current/setup.html.