Table of Contents

Adding an Elasticsearch cluster node

Important

Elasticsearch is only supported up to version 6.8, which is no longer supported by Elastic. We therefore recommend using Storage as a Service instead, or if you do want to continue using self-hosted storage, using OpenSearch.

Prerequisites

  • On the node to be added, make sure the data folder is empty. In the elasticsearch.yml file, you can find the data folder in the path.data setting.
  • Make sure the Elasticsearch service is stopped. Data will be added to the data folder as soon as the service is started again.
  • On the node to be added, make sure port 9200 and 9300 are open. Note that these are the default ports and that alternative ports can be defined in the elasticsearch.yml file.
  • If you intend to add the node to a single-node cluster, make sure that node also has ports 9200 and 9300 open. Note that these are the default ports and that alternative ports can be defined in the elasticsearch.yml file.
  • Make sure the network.host is bound to an IP address that is accessible to the other nodes in the cluster (other than 127.0.0.1).
  • Make sure the network.publish_host is bound to an IP address that is accessible to the other nodes in the cluster (other than 127.0.0.1). If this setting is not explicitly defined, its value will be inherited from network.host.

Adding the node

  1. On the node to be added to the cluster, open the elasticsearch.yml file and configure the following settings:

    Setting Description
    cluster.name:NAME This name should be identical to the cluster name specified on the other nodes.
    network.host:IP Make sure the network.host is bound to an IP address that is accessible to the other nodes in the cluster (other than 127.0.0.1).
    Network.publish_host:IP Make sure the network.publish_host is bound to an IP address that is accessible to the other nodes in the cluster (other than 127.0.0.1). If this setting is not explicitly defined, its value will be inherited from network.host.
    discovery.zen.minimum_master_nodes:X This value should be identical to the discovery.zen.minimum_master_nodes values specified on the other nodes. For more information on this setting, see Configuring the master nodes.
    node.master:false In most cases, this setting will be set to false. For more information on this setting, see Configuring the master nodes.
    discovery.zen.ping.unicast.hosts:[“IP1”, “IP2”, ...] This setting should contain the IP addresses of the (master) nodes in the cluster.

  2. Start the node to have it added to the cluster.

Tip

See also: Standalone Elasticsearch Cluster Installer

Confirming that the node has been added

To get confirmation that the node has successfully been added to the cluster, execute the _cat/nodes query on any of the other nodes in the cluster.

Adding a potential master node

In some cases, (multiple) potential master nodes should be added to the cluster.

On those potential master nodes, node.master must be set to true and discovery.zen.minimum_master_nodes must contain the minimum number of potential master nodes in the cluster. Note that the latter should also be adjusted on the existing potential master nodes.

After configuring the necessary settings on the potential master node to be added, restart the existing (potential) master nodes, and then start the potential master node to be added.

For more information on master nodes, see Configuring the master nodes.

Propose changes