Lab 8.1: Multi cluster operations

Objective:

In this lab, you will set up a remote cluster for cross-cluster search. You will also configure cross-cluster replication.

Click the Kibana 2 button in Strigo to open Kibana, which is connected to a one-node cluster named cluster2 that is already up and running. This node is accessible at the address node5:9304. Log in using training as the username and nonprodpwd for the password.
Using Console, run a match_all query on the blogs index on cluster2. You will see that there are seven blogs, and notice they were published over three days: April 27-29, 2021.
```
GET blogs/_search
```
Go back to Kibana 1 and register cluster2 as a remote cluster to cluster1.
Solution

Go to Stack Management, then select Remote Clusters. Add a new remote cluster with the following properties:
- Name: cluster2
- Seed nodes: node5:9304
- Node connections: 1
Click Save and make sure that the cluster is connected.
Write a query in Console of Kibana 1 that returns the seven blogs on cluster2.
Solution

A simple match_all query will do the trick - just make sure the index name is cluster2:blogs:
```
GET cluster2:blogs/_search
```
Write a query that searches the blogs indices on cluster1 and cluster2. Get the blogs containing the phrase "kibana query language" in the content field. You should get 24 hits.
Solution
```
GET blogs,cluster2:blogs/_search
{
  "query": {
    "match_phrase": {
      "content": "kibana query language"
    }
  }
}
```
Next, you will configure cluster1 to replicate the blogs index from cluster2 (which currently only has seven documents). From Kibana1, go to the Cross-Cluster Replication page in Stack Management and select Create a follower index.
Configure CCR so that the blogs index on cluster2 is replicated to cluster1 in a new index named replicated_blogs.
Solution

You can also complete this task by running the following command in Console of Kibana1:
```
PUT /replicated_blogs/_ccr/follow
{
  "remote_cluster": "cluster2",
  "leader_index": "blogs"
}
```
Your new follower index is paused initially. Start the replication process now.

Solution

You can start replication either by clicking the Manage button in the bottom-right corner of the screen you are viewing. If you already closed that window, the same options appear by clicking the three dots under Actions on the Cross-Cluster Replication UI page.
Run the following query in Console of cluster1. You should get seven hits:
```
GET /replicated_blogs/_count
```
Let's change the leader index and see what happens. From Console of cluster2 (which is the Kibana 2 button in Strigo), run the following query - which adds a new field to each document in the blogs index:
```
POST blogs/_update_by_query
{
  "script": {
    "source": "ctx._source.version = 2",
    "lang": "painless"
  }
}
```
Go back to Console of cluster1 and view the documents in replicated_blogs. They should each have a new field named version that was added on cluster2:
```
GET replicated_blogs/_search
```

Summary:

In this lab, you wrote a search across indices on multiple clusters. You implemented cross-cluster replication by replicating the blogs index of cluster2 to a new index on cluster1 named replicated_blogs.