Adding a New Node to an Existing Cluster

As determined by your business requirements for high availability and disaster recovery, you will need to determine the number of nodes appropriate for your installation. In addition to high availability and disaster recovery, system CPU and memory limitations may require addition of nodes to meet performance requirements. This section discusses how to add a node to your existing cluster.

To add a new node to an existing Elasticsearch cluster, you should complete the following steps:

  1. Install Elasticsearch on a separate machine using the instructions provided in the PeopleSoft Deployment Packages for Elasticsearch Installation (Doc ID 2205540.2) on My Oracle Support. for your operating system.

    • For the cluster name prompt, ensure that you enter the name of the existing cluster to which you want to add this node.

    • For the minimum number of master nodes prompt, before you enter a value consider the parameter description that is discussed in an earlier section.

      See Minimum Number of Master Nodes.

    • For the host names of existing nodes, ensure that you enter the host names of the existing nodes so that the new node will discover the other nodes in the cluster.

    Complete the Elasticsearch installation by following the steps described in PeopleSoft Deployment Packages for Elasticsearch Installation. After completing the installation, verify the cluster and node setup.

  2. Open the elasticsearch.yml configuration file of the existing nodes in the cluster and edit the following parameters.

    Field or Control

    Definition

    discovery.zen.minimum_master_nodes

    Enter the value that you provided in Step 1.

    discovery.zen.ping.unicast.hosts

    Add the host of the new node to the existing list of hosts. Elasticsearch nodes will find each other via unicast. Ensure that you enter the address in the correct format: ["host1", "host2"].

    See List of Host Names where Elasticsearch will be Installed or is Installed.

  3. The elasticsearch.yml files of the existing nodes is updated with the host of the new node so that, in the future, whenever any of the nodes is restarted, the restarted node will be able to discover other nodes in the cluster mentioned in the discovery.zen.ping.unicast.hosts parameter.

    Note: An .yml file uses space indentation, so ensure that you edit the file properly and save it.

  4. On PeopleSoft Search Framework, add the details of the new node to the Elasticsearch search instance on the Search Instance Properties page.

    See Working With Search Instances.