Deployments for Kibana instances

Hello and welcome back!

In this article, we will guide you through deploying a Kibana pod on your Kubernetes cluster. You will learn how to verify your cluster’s status, review the necessary deployment and service manifests, and confirm that Kibana connects effectively with Elasticsearch.

1. Verifying the Cluster Status

Before proceeding, ensure your cluster is healthy by checking the status of your pods. Run the following command:

kubectl get pods

You should see output similar to this:

NAME                                READY   STATUS    RESTARTS   AGE
elasticsearch-0                     1/1     Running   0          2m50s

Next, verify that the required manifest files are available in your current working directory. List the files by executing:

ls -lrt

A sample output might be:

total 20
-rw-r--r--  1 root root   184 Aug  6 14:18 kibana-service.yaml
-rw-r--r--  1 root root   354 Aug  6 14:18 kibana-deployment.yaml
-rw-r--r--  1 root root  1195 Aug  6 14:18 es-statefulset.yaml
-rw-r--r--  1 root root   209 Aug  6 14:18 es-service.yaml
-rw-r--r--  1 root root   185 Aug  6 14:18 es-volume.yaml

2. Reviewing the Kibana Deployment Manifest

The kibana-deployment.yaml file defines how Kubernetes should deploy the Kibana pod. Open the file to review its contents:

cat kibana-deployment.yaml

The manifest content is as follows:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kibana
  namespace: efk
spec:
  replicas: 1
  selector:
    matchLabels:
      app: kibana
  template:
    metadata:
      labels:
        app: kibana
    spec:
      containers:
      - name: kibana
        image: docker.elastic.co/kibana/kibana:8.13.0
        ports:
        - containerPort: 5601

This manifest specifies:

• Resource Kind: A Deployment used to manage the lifecycle of the Kibana pod.
• Namespace: The efk namespace.
• Replica Count: A single instance (replica) of Kibana.
• Container Configuration: Utilizes the official Kibana Docker image from Elastic and exposes port 5601.

3. Reviewing the Kibana Service Manifest

The kibana-service.yaml file exposes Kibana using a NodePort. Open and inspect the service manifest:

cat kibana-service.yaml

The manifest content should appear as:

apiVersion: v1
kind: Service
metadata:
  name: kibana
  namespace: efk
spec:
  type: NodePort
  selector:
    app: kibana
  ports:
    - protocol: TCP
      port: 5601
      nodePort: 30601

This service configuration:

• Uses the NodePort type to make Kibana accessible externally.
• Routes traffic from the cluster’s internal port 5601 to the node’s port 30601.
• Targets pods labeled with app: kibana in the efk namespace.

4. Deploying Kibana

Step 1: Apply the Deployment Manifest

Deploy the Kibana pod by applying the deployment manifest:

kubectl apply -f kibana-deployment.yaml

A successful operation will confirm with a message similar to:

deployment.apps/kibana created

Step 2: Apply the Service Manifest

Next, expose the Kibana pod by applying the service manifest:

kubectl apply -f kibana-service.yaml

You should see an output like:

service/kibana created

Step 3: Verify the Pod Status

Clear your terminal and run the following command to check the status of the pods:

kubectl get pods

The output may resemble:

NAME                                READY   STATUS              RESTARTS   AGE
elasticsearch-0                     1/1     Running             0          4m47s
kibana-5bf7c766b4-pgdz              0/1     ContainerCreating   0          14s

Wait a minute or two for the Kibana pod to transition to the Running state. To monitor the progress in real-time, use:

kubectl get pods -w

Once the pod shows as running, proceed to verify your services:

kubectl get svc

A sample output might be:

NAME            TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)                         AGE
elasticsearch   NodePort    10.96.188.157   <none>        9200:32000/TCP,9300:30300/TCP     5m30s
kibana          NodePort    10.103.134.155  <none>        5601:30601/TCP                  55s

5. Connecting to the Kibana UI

Follow these steps to access the Kibana user interface:

1. Identify the NodePort: In our example, the NodePort is 30601.

2. Access through Browser: Open your web browser and enter the following URL, replacing <NODE_IP> with the IP address of your node:

   http://<NODE_IP>:30601

Once the Kibana UI loads, click on "Explore on my own" (or a similar option) and proceed to the "Dev Tools" section. This interactive console allows you to run queries against the connected Elasticsearch cluster.

Testing Elasticsearch Connectivity

Use the Dev Tools console in Kibana to execute the following commands:

1. Run a Search Query:

   GET _search
   {
     "query": {
       "match_all": {}
     }
   }
   

2. Check Cluster Health:

   Clear the previous output and run:

   GET /_cluster/health
   

A successful response will be similar to:

{
  "cluster_name": "docker-cluster",
  "status": "green",
  "timed_out": false,
  "number_of_nodes": 1,
  "number_of_data_nodes": 1,
  "active_primary_shards": 28,
  "active_shards": 28,
  "relocating_shards": 0,
  "initializing_shards": 0,
  "unassigned_shards": 0,
  "delayed_unassigned_shards": 0,
  "number_of_pending_tasks": 0,
  "number_of_in_flight_fetch": 0,
  "task_max_waiting_in_queue_millis": 0,
  "active_shards_percent_as_number": 100
}

A status of green means your Elasticsearch cluster is healthy and that Kibana is properly connected.

Conclusion

In this guide, you learned how to deploy and expose Kibana on your Kubernetes cluster within the efk namespace. By carefully following the verification, deployment, and connectivity tests, you can replicate this setup in your environment—whether on-premises or in the cloud.

Happy deploying, and see you in the next article!

Additional Resources

• Kubernetes Documentation
• Kibana Documentation
• Elasticsearch Documentation

If you have any questions, feel free to consult the linked resources or reach out for further assistance.