Understanding High Availability Feature in PeopleSoft Search Framework
Elasticsearch uses distributed architecture and is built to be always available. Elasticsearch uses clusters and nodes to provide high availability.
Oracle PeopleSoft’s recommended approach to high availability is to deploy an Elasticsearch cluster with at least 3 nodes on machines that are not on the same physical server and by following the configuration guidelines provided by PeopleSoft Search Framework. The high availability feature of Elasticsearch enables PeopleSoft Search Framework to provide a fail-over mechanism.
PeopleSoft Search Framework uses Elasticsearch search engine to provide search functionality. Elasticsearch search engine is distributed by nature, that is, it knows how to manage multiple nodes and manages failover through primary shards and replicas. To enable users to take advantage of this fail-over mechanism, PeopleSoft Search Framework provides you the framework to configure multiple nodes in a search instance of your PeopleSoft implementation. Using the constructs of cluster and node of Elasticsearch, a search instance in PeopleSoft represents an Elasticsearch cluster that can contain multiple nodes.
See Implementing Fail-over Mechanism Using Multiple Nodes Feature of Elasticsearch.
In addition to the built-in fail-over mechanism of Elasticsearch, PeopleSoft Search Framework provides you another fail-over mechanism by using its feature of configuring multiple search instances for the Elasticsearch search engine. Implementing this feature means that you are creating multiple Elasticsearch clusters.