Helm charts to install the Elastic Stack using minikube

On the previous post we used the minikube wrapper kube.sh to install ECK – Elastic Cloud on Kubernetes. ECK uses an operator that was created by Elastic to deploy and orchestrate the Elastic Stack in kubernetes. ECK is not the only way to install elasticsearch in kubernetes, you can also use Helm charts to install the Stack.

What is the difference between an operator and helm charts?

Helm

Helm is a package management system for kubernetes. The packaging format is called charts. In OS terms its like rpm or deb packages. An application is packed into a package that can be installed using a helm install command. It is used to get an application up and running quickly. You can run packages with minimal changes or even no changes at all and you can quickly repeat the application on multiple environments quickly.

Kubernetes operators

Operators also package applications into an easy to deploy formats but it also uses custom resources to include a glot of complex configuration data within the package. With operators you can deploy a stateful application in a automatic way and deploy an application across a cluster that is configured in a particular way to achieve HA. You can also automate other tasks via orchestration via operators.

When to use helm vs operators?

Simple application install – any type of simple application install wtihout much customization, Helm is a better and easier choice
Customizations? – If you are configuring application with a lot of configurations or orchestrations, operator would be better
Iteration/maturity – If your application is mature and went through various iterations and you know how it works in the kubernetes environment and want to have some automation as next steps, operator would be better

Even with the differences both helm and operator has its uses. In this article we will install helm and configure Elasticsearch and kibana using helm charts.

Install Helm

Helm provides a super easy way to install the helm binary for any OS.

$ curl https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3 | bash

Lets install kube.sh and start our minikube environment

$ curl -fsSL https://raw.githubusercontent.com/jlim0930/scripts/master/kube.sh -o kube.sh
$ chmod +x kube.sh
$ ./kube.sh start
bash: warning: setlocale: LC_ALL: cannot change locale (C.UTF-8)
[DEBUG] minikube found.
[DEBUG] kubectl found.
[DEBUG] build minikube
[DEBUG] CPU will be set to 4 cores
❗  These changes will take effect upon a minikube delete and then a minikube start
[DEBUG] MEM will be set to 16005mb
❗  These changes will take effect upon a minikube delete and then a minikube start
😄  minikube v1.19.0 on Centos 7.9.2009
🎉  minikube 1.20.0 is available! Download it: https://github.com/kubernetes/minikube/releases/tag/v1.20.0
💡  To disable this notice, run: 'minikube config set WantUpdateNotification false'

✨  Using the docker driver based on user configuration
👍  Starting control plane node minikube in cluster minikube
💾  Downloading Kubernetes v1.20.2 preload ...
    > preloaded-images-k8s-v10-v1...: 491.71 MiB / 491.71 MiB  100.00% 10.99 Mi
🔥  Creating docker container (CPUs=4, Memory=16005MB) ...
🐳  Preparing Kubernetes v1.20.2 on Docker 20.10.5 ...
    ▪ Generating certificates and keys ...
    ▪ Booting up control plane ...
    ▪ Configuring RBAC rules ...
🔎  Verifying Kubernetes components...
    ▪ Using image gcr.io/k8s-minikube/storage-provisioner:v5
🌟  Enabled addons: storage-provisioner, default-storageclass
🏄  Done! kubectl is now configured to use "minikube" cluster and "default" namespace by default
    ▪ Using image metallb/speaker:v0.8.2
    ▪ Using image metallb/controller:v0.8.2
🌟  The 'metallb' addon is enabled
[DEBUG] minikube IP is: 192.168.49.2
[DEBUG] LoadBalancer Pool: 192.168.49.150 - 192.168.49.175

Please note that we have a LoadBalancer Pool of 192.168.49.150-192.168.49.175. The range might be different on your machine but you should be able to browse or access these IP’s from the same server/workstation.

Now we have everything we need and a running minikube environment to get going.

Elastic’s Helm charts are located: https://github.com/elastic/helm-charts
Elasticsearch/Kibana/Filebeat/Metricbeat was GA’ed since 7.7 and currently as of today logstash and APM server is still in BETA.

Although you can install ECK operator using Helm charts as seen on https://www.elastic.co/guide/en/cloud-on-k8s/current/k8s-install-helm.html we will install Elasticsearch and Kibana as applications without the operator.

Lets get started

Add the elastic repo

$ helm repo add elastic https://helm.elastic.co
"elastic" has been added to your repositories

To see all packages in the elastic repository

$ helm search repo elastic
NAME                        CHART VERSION   APP VERSION DESCRIPTION
elastic/elasticsearch       7.12.1          7.12.1      Official Elastic helm chart for Elasticsearch
elastic/apm-server          7.12.1          7.12.1      Official Elastic helm chart for Elastic APM Server
elastic/eck-operator        1.5.0           1.5.0       A Helm chart for deploying the Elastic Cloud on...
elastic/eck-operator-crds   1.5.0           1.5.0       A Helm chart for installing the ECK operator Cu...
elastic/filebeat            7.12.1          7.12.1      Official Elastic helm chart for Filebeat
elastic/kibana              7.12.1          7.12.1      Official Elastic helm chart for Kibana
elastic/logstash            7.12.1          7.12.1      Official Elastic helm chart for Logstash
elastic/metricbeat          7.12.1          7.12.1      Official Elastic helm chart for Metricbeat

You can add the -l switch to get the long listing that shows all the available versions helm search repo elastic -l

Install elasticsearch

There are multiple ways to install a package

You can just run helm install elasticsearch elastic/elasticsearch and it will install elasticsearch using the default values set in the helm chart.

You can download all the default values for the chart and edit it to customize the chart and install using the customized values

$ helm show values elastic/elasticserach | tee -a values.yaml
$ vi values.yaml # Edit for your needs
$ helm install elasticsearch elastic/elasticsearch -f ./values.yaml

You can install using just --set to set the values of customization inline
```
$ helm install elasticsearch elastic/elasticsearch --set xxxxxxxxxxxx
```
NOTES – all custom values and their meanings can be found on https://github.com/elastic/helm-charts/tree/master/elasticsearch Also it would be good to look through the values.yaml to see what you can set.

For our purposes I will install it using inline options

$ helm install elasticsearch elastic/elasticsearch --set service.type=LoadBalancer,antiAffinity="soft",replicas=3,imageTag=7.12.1
NAME: elasticsearch
LAST DEPLOYED: Sat May 15 09:34:35 2021
NAMESPACE: default
STATUS: deployed
REVISION: 1
NOTES:
1. Watch all cluster members come up.
  $ kubectl get pods --namespace=default -l app=elasticsearch-master -w
2. Test cluster health using Helm test.
  $ helm test elasticsearch

$ kubectl get pods --namespace=default -l app=elasticsearch-master -w
NAME                     READY   STATUS     RESTARTS   AGE
elasticsearch-master-0   0/1     Init:0/1   0          7s
elasticsearch-master-1   0/1     Init:0/1   0          7s
elasticsearch-master-2   0/1     Init:0/1   0          7s
elasticsearch-master-1   0/1     PodInitializing   0          53s
elasticsearch-master-2   0/1     PodInitializing   0          53s
elasticsearch-master-0   0/1     PodInitializing   0          53s
elasticsearch-master-0   0/1     Running           0          54s
elasticsearch-master-2   0/1     Running           0          54s
elasticsearch-master-1   0/1     Running           0          55s
elasticsearch-master-1   1/1     Running           0          118s
elasticsearch-master-0   1/1     Running           0          2m2s
elasticsearch-master-2   1/1     Running           0          2m4s

$ helm test elasticsearch
NAME: elasticsearch
LAST DEPLOYED: Sat May 15 09:34:35 2021
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE:     elasticsearch-gtuee-test
Last Started:   Sat May 15 09:36:57 2021
Last Completed: Sat May 15 09:36:59 2021
Phase:          Succeeded
NOTES:
1. Watch all cluster members come up.
  $ kubectl get pods --namespace=default -l app=elasticsearch-master -w
2. Test cluster health using Helm test.
  $ helm test elasticsearch

$ kubectl get pod,svc,pv
NAME                         READY   STATUS    RESTARTS   AGE
pod/elasticsearch-master-0   1/1     Running   0          3m27s
pod/elasticsearch-master-1   1/1     Running   0          3m27s
pod/elasticsearch-master-2   1/1     Running   0          3m27s

NAME                                    TYPE           CLUSTER-IP       EXTERNAL-IP      PORT(S)                         AGE
service/elasticsearch-master            LoadBalancer   10.102.211.193   192.168.49.150   9200:30416/TCP,9300:31121/TCP   3m27s
service/elasticsearch-master-headless   ClusterIP      None             <none>           9200/TCP,9300/TCP               3m27s
service/kubernetes                      ClusterIP      10.96.0.1        <none>           443/TCP                         7m37s

NAME                                                        CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                                                 STORAGECLASS   REASON   AGE
persistentvolume/pvc-2cc64f66-2967-4734-92f4-6c428effd516   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-0   standard                3m27s
persistentvolume/pvc-61e9fa07-1ae9-4741-b835-e0b0877476ba   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-2   standard                3m27s
persistentvolume/pvc-9481c420-786a-4edc-a200-67ea5a043fb7   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-1   standard                3m27s

When I installed elasticsearch I passed a few settings service.type=LoadBalancer,antiAffinity="soft",replicas=3,imageTag=7.12.1 These settings can be found on the default values.yaml. I just passed the settings that I wanted to update and change from the default.

Setting	Changed
service.type	default is `ClusterIP` but I wanted to make it reachable and used a `LoadBalancer`
antiAffinity	default is `hard` however since I am using minikube I need to set it as `soft`
replicas	default is `1` I wanted to have 3 nodes
imageTag	You can set this to any supported version of elastic

Now that our cluster is up lets access it. As you scan also see from the above our elasticsearch-master service has an external IP of 192.168.49.150 so we’ll use that to access it

$ curl 192.168.49.150:9200
{
  "name" : "elasticsearch-master-0",
  "cluster_name" : "elasticsearch",
  "cluster_uuid" : "MNEDcCWwQMOLZrYmQDoIvw",
  "version" : {
    "number" : "7.12.1",
    "build_flavor" : "default",
    "build_type" : "docker",
    "build_hash" : "3186837139b9c6b6d23c3200870651f10d3343b7",
    "build_date" : "2021-04-20T20:56:39.040728659Z",
    "build_snapshot" : false,
    "lucene_version" : "8.8.0",
    "minimum_wire_compatibility_version" : "6.8.0",
    "minimum_index_compatibility_version" : "6.0.0-beta1"
  },
  "tagline" : "You Know, for Search"
}

$ curl 192.168.49.150:9200/_cat/nodes
172.17.0.6 41 62 10 0.35 0.45 0.56 cdfhilmrstw - elasticsearch-master-2
172.17.0.4 27 62 10 0.35 0.45 0.56 cdfhilmrstw - elasticsearch-master-0
172.17.0.5 46 62 10 0.35 0.45 0.56 cdfhilmrstw * elasticsearch-master-1

The defaults does not setup any type of security and we can get into that later on.

Install kibana

Again you can get the defaults by running

$ helm show values elastic/kibana | tee -a values.yaml

$ vi values.yaml

For this example I will just run a simple kibana install via inline with minimal settings

$ helm install kibana elastic/kibana --set imageTag=7.12.1,service.type=LoadBalancer
NAME: kibana
LAST DEPLOYED: Sat May 15 09:52:24 2021
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None

$ kubectl get pod -w
NAME                             READY   STATUS              RESTARTS   AGE
elasticsearch-master-0           1/1     Running             0          18m
elasticsearch-master-1           1/1     Running             0          18m
elasticsearch-master-2           1/1     Running             0          18m
kibana-kibana-5fb66568d8-vmr2f   0/1     ContainerCreating   0          40s
kibana-kibana-5fb66568d8-vmr2f   0/1     Running             0          54s
kibana-kibana-5fb66568d8-vmr2f   1/1     Running             0          103s

$ kubectl get pod,svc,pv
NAME                                 READY   STATUS    RESTARTS   AGE
pod/elasticsearch-master-0           1/1     Running   0          21m
pod/elasticsearch-master-1           1/1     Running   0          21m
pod/elasticsearch-master-2           1/1     Running   0          21m
pod/kibana-kibana-5fb66568d8-vmr2f   1/1     Running   0          3m42s

NAME                                    TYPE           CLUSTER-IP       EXTERNAL-IP      PORT(S)                         AGE
service/elasticsearch-master            LoadBalancer   10.102.211.193   192.168.49.150   9200:30416/TCP,9300:31121/TCP   21m
service/elasticsearch-master-headless   ClusterIP      None             <none>           9200/TCP,9300/TCP               21m
service/kibana-kibana                   LoadBalancer   10.99.43.212     192.168.49.151   5601:32143/TCP                  3m42s
service/kubernetes                      ClusterIP      10.96.0.1        <none>           443/TCP                         25m

NAME                                                        CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM                                                 STORAGECLASS   REASON   AGE
persistentvolume/pvc-2cc64f66-2967-4734-92f4-6c428effd516   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-0   standard                21m
persistentvolume/pvc-61e9fa07-1ae9-4741-b835-e0b0877476ba   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-2   standard                21m
persistentvolume/pvc-9481c420-786a-4edc-a200-67ea5a043fb7   30Gi       RWO            Delete           Bound    default/elasticsearch-master-elasticsearch-master-1   standard                21m

Looks like our kibana is running on 192.168.49.151

Again you can edit values and install using -f values.yaml instead of inline.

I will be updating this article with SSL configuration, Security settings, and various use cases. If you have any usecase you want to suggest please let me know.