Tutorial: Set up cross-cluster replication

edit

Use this guide to set up cross-cluster replication (CCR) between clusters in two datacenters. Replicating your data across datacenters provides several benefits:

  • Brings data closer to your users or application server to reduce latency and response time
  • Provides your mission-critical applications with the tolerance to withstand datacenter or region outages

In this guide, you’ll learn how to:

  • Configure a remote cluster with a leader index
  • Create a follower index on a local cluster
  • Create an auto-follow pattern to automatically follow time series indices that are periodically created in a remote cluster

You can manually create follower indices to replicate specific indices on a remote cluster, or configure auto-follow patterns to replicate rolling time series indices.

If you want to replicate data across clusters in the cloud, you can configure remote clusters on Elasticsearch Service. Then, you can search across clusters and set up cross-cluster replication.

Prerequisites

edit

To complete this tutorial, you need:

  • The manage cluster privilege on the local cluster.
  • A license on both clusters that includes cross-cluster replication. Activate a free 30-day trial.
  • An index on the remote cluster that contains the data you want to replicate. This tutorial uses the sample eCommerce orders data set. Load sample data.
  • In the local cluster, all nodes with the master node role must also have the remote_cluster_client role. The local cluster must also have at least one node with both a data role and the remote_cluster_client role. Individual tasks for coordinating replication scale based on the number of data nodes with the remote_cluster_client role in the local cluster.

Connect to a remote cluster

edit

To replicate an index on a remote cluster (Cluster A) to a local cluster (Cluster B), you configure Cluster A as a remote on Cluster B.

ClusterA contains the leader index and ClusterB contains the follower index

To configure a remote cluster from Stack Management in Kibana:

  1. Select Remote Clusters from the side navigation.
  2. Specify the Elasticsearch endpoint URL, or the IP address or host name of the remote cluster (ClusterA) followed by the transport port (defaults to 9300). For example, cluster.es.eastus2.staging.azure.foundit.no:9400 or 192.168.1.1:9300.
API example

You can also use the cluster update settings API to add a remote cluster:

PUT /_cluster/settings
{
  "persistent" : {
    "cluster" : {
      "remote" : {
        "leader" : {
          "seeds" : [
            "127.0.0.1:9300" 
          ]
        }
      }
    }
  }
}

Specifies the hostname and transport port of a seed node in the remote cluster.

You can verify that the local cluster is successfully connected to the remote cluster.

GET /_remote/info

The API response indicates that the local cluster is connected to the remote cluster with cluster alias leader.

{
  "leader" : {
    "seeds" : [
      "127.0.0.1:9300"
    ],
    "connected" : true,
    "num_nodes_connected" : 1, 
    "max_connections_per_cluster" : 3,
    "initial_connect_timeout" : "30s",
    "skip_unavailable" : false,
    "mode" : "sniff"
  }
}

The number of nodes in the remote cluster the local cluster is connected to.

Configure privileges for cross-cluster replication

edit

The cross-cluster replication user requires different cluster and index privileges on the remote cluster and local cluster. Use the following requests to create separate roles on the local and remote clusters, and then create a user with the required roles.

Remote cluster
edit

On the remote cluster that contains the leader index, the cross-cluster replication role requires the read_ccr cluster privilege, and monitor and read privileges on the leader index.

If requests will be issued on behalf of other users, then the the authenticating user must have the run_as privilege on the remote cluster.

The following request creates a remote-replication role on the remote cluster:

POST /_security/role/remote-replication
{
  "cluster": [
    "read_ccr"
  ],
  "indices": [
    {
      "names": [
        "leader-index-name"
      ],
      "privileges": [
        "monitor",
        "read"
      ]
    }
  ]
}
Local cluster
edit

On the local cluster that contains the follower index, the remote-replication role requires the manage_ccr cluster privilege, and monitor, read, write, and manage_follow_index privileges on the follower index.

The following request creates a remote-replication role on the local cluster:

POST /_security/role/remote-replication
{
  "cluster": [
    "manage_ccr"
  ],
  "indices": [
    {
      "names": [
        "follower-index-name"
      ],
      "privileges": [
        "monitor",
        "read",
        "write",
        "manage_follow_index"
      ]
    }
  ]
}

After creating the remote-replication role on each cluster, use the create or update users API to create a user on the local cluster cluster and assign the remote-replication role. For example, the following request assigns the remote-replication role to a user named cross-cluster-user:

POST /_security/user/cross-cluster-user
{
  "password" : "l0ng-r4nd0m-p@ssw0rd",
  "roles" : [ "remote-replication" ]
}

You only need to create this user on the local cluster.

Create a follower index to replicate a specific index

edit

When you create a follower index, you reference the remote cluster and the leader index in your remote cluster.

To create a follower index from Stack Management in Kibana:

  1. Select Cross-Cluster Replication in the side navigation and choose the Follower Indices tab.
  2. Choose the cluster (ClusterA) containing the leader index you want to replicate.
  3. Enter the name of the leader index, which is kibana_sample_data_ecommerce if you are following the tutorial.
  4. Enter a name for your follower index, such as follower-kibana-sample-data.

Elasticsearch initializes the follower using the remote recovery process, which transfers the existing Lucene segment files from the leader index to the follower index. The index status changes to Paused. When the remote recovery process is complete, the index following begins and the status changes to Active.

When you index documents into your leader index, Elasticsearch replicates the documents in the follower index.

The Cross-Cluster Replication page in Kibana
API example

You can also use the create follower API to create follower indices. When you create a follower index, you must reference the remote cluster and the leader index that you created in the remote cluster.

When initiating the follower request, the response returns before the remote recovery process completes. To wait for the process to complete, add the wait_for_active_shards parameter to your request.

PUT /server-metrics-follower/_ccr/follow?wait_for_active_shards=1
{
  "remote_cluster" : "leader",
  "leader_index" : "server-metrics"
}

Use the get follower stats API to inspect the status of replication.

Create an auto-follow pattern to replicate time series indices

edit

You use auto-follow patterns to automatically create new followers for rolling time series indices. Whenever the name of a new index on the remote cluster matches the auto-follow pattern, a corresponding follower index is added to the local cluster. Note that only indices created on the remote cluster after the auto-follow pattern is created will be auto-followed: existing indices on the remote cluster are ignored even if they match the pattern.

An auto-follow pattern specifies the remote cluster you want to replicate from, and one or more index patterns that specify the rolling time series indices you want to replicate.

To create an auto-follow pattern from Stack Management in Kibana:

  1. Select Cross Cluster Replication in the side navigation and choose the Auto-follow patterns tab.
  2. Enter a name for the auto-follow pattern, such as beats.
  3. Choose the remote cluster that contains the index you want to replicate, which in the example scenario is Cluster A.
  4. Enter one or more index patterns that identify the indices you want to replicate from the remote cluster. For example, enter metricbeat-* packetbeat-* to automatically create followers for Metricbeat and Packetbeat indices.
  5. Enter follower- as the prefix to apply to the names of the follower indices so you can more easily identify replicated indices.

As new indices matching these patterns are created on the remote, Elasticsearch automatically replicates them to local follower indices.

The Auto-follow patterns page in Kibana
API example

Use the create auto-follow pattern API to configure auto-follow patterns.

PUT /_ccr/auto_follow/beats
{
  "remote_cluster" : "leader",
  "leader_index_patterns" :
  [
    "metricbeat-*", 
    "packetbeat-*" 
  ],
  "follow_index_pattern" : "{{leader_index}}-copy" 
}

Automatically follow new Metricbeat indices.

Automatically follow new Packetbeat indices.

The name of the follower index is derived from the name of the leader index by adding the suffix -copy to the name of the leader index.