https://medium.com/avmconsulting-blog/how-to-deploy-an-efk-stack-to-kubernetes-ebc1b539d063

https://www.digitalocean.com/community/tutorials/how-to-set-up-an-elasticsearch-fluentd-and-kibana-efk-logging-stack-on-kubernetes


01. 개요

02. 사전 조건

  • Kubernetes 1.16+
  • Helm 3.x

03. 설치

03-01. Namespace 생성

  • EFK Stack을 Deploy할 namespace를 생성합니다.
sansae@win10pro-worksp:/workspaces$ kubectl create ns monitor-efk

03-02. StorageClass 확인

  • Elasticsearch의 PersistanceVolume을 사용하기 위하여, StorageClass를 확인 합니다.
  • Elasticsearch는 AzureDisk로 Volume을 생성해야 합니다. 따라서 'managed-premium'을 사용할 것입니다.
sansae@win10pro-worksp:/workspaces$ kubectl get sc
NAME                PROVISIONER                RECLAIMPOLICY   VOLUMEBINDINGMODE      ALLOWVOLUMEEXPANSION   AGE
azurefile           kubernetes.io/azure-file   Delete          Immediate              true                   63d
azurefile-premium   kubernetes.io/azure-file   Delete          Immediate              true                   63d
default (default)   kubernetes.io/azure-disk   Delete          Immediate              true                   63d
managed             kubernetes.io/azure-disk   Delete          WaitForFirstConsumer   true                   28d
managed-premium     kubernetes.io/azure-disk   Delete          Immediate              true                   63d

03-03. Elasticsearch

  • Elasticsearch를 Stable버전으로 명세 합니다.
  • Volume을 영구 저장하기 위하여 PersistanceVolume를 Azure Disk로 할당 받아야 합니다. 따라서, StorageClassName을 'managed-premium'으로 설정합니다.
sansae@win10pro-worksp:/workspaces$ git clone https://github.com/elastic/helm-charts.git
sansae@win10pro-worksp:/workspaces$ mv helm-charts elastic
sansae@win10pro-worksp:/workspaces$ cd elastic
sansae@win10pro-worksp:/workspaces/elastic$ 
sansae@win10pro-worksp:/workspaces/elastic$ vi elasticsearch/values.yaml

=========================================================
imageTag: "7.12.0"
----------------------------------------
volumeClaimTemplate:
  storageClassName: managed-premium
  accessModes: [ "ReadWriteOnce" ]
  resources:
    requests:
      storage: 30Gi
=========================================================

sansae@win10pro-worksp:/workspaces/elastic$ helm install elasticsearch/ --generate-name -n monitor-efk
NAME: elasticsearch-1617184828
LAST DEPLOYED: Wed Mar 31 19:00:30 2021
NAMESPACE: monitor-efk
STATUS: deployed
REVISION: 1
NOTES:
1. Watch all cluster members come up.
  $ kubectl get pods --namespace=monitor-efk -l app=elasticsearch-master -w
2. Test cluster health using Helm test.
  $ helm --namespace=monitor-efk test elasticsearch-1617184828


sansae@win10pro-worksp:/workspaces/elastic$ kubectl get pod -n monitor-efk
NAME                       READY   STATUS    RESTARTS   AGE
elasticsearch-master-0     1/1     Running   0          4m20s
elasticsearch-master-1     1/1     Running   0          4m20s
elasticsearch-master-2     1/1     Running   0          4m20s

03-03. Fluentd

  • Elasticsearch ServiceName 'elasticsearch-master'로 설정합니다.
  • Elasticsearch Port: 9200 으로 설정합니다.
sansae@win10pro-worksp:/workspaces$ git clone https://github.com/fluent/helm-charts.git
sansae@win10pro-worksp:/workspaces$ mv helm-charts fluent
sansae@win10pro-worksp:/workspaces$ cd fluent
sansae@win10pro-worksp:/workspaces/fluent$ cd charts
sansae@win10pro-worksp:/workspaces/fluent/charts$ vi fluentd/values.yaml

=========================================================
env:
#- name: "FLUENTD_CONF"
#  value: "../../etc/fluent/fluent.conf"
  - name: FLUENT_ELASTICSEARCH_HOST
    value: "elasticsearch-master"
  - name: FLUENT_ELASTICSEARCH_PORT
    value: "9200"
=========================================================

sansae@win10pro-worksp:/workspaces/fluent/charts$ helm install fluentd --generate-name -n monitor-efk
NAME: fluentd-1617186347
LAST DEPLOYED: Wed Mar 31 19:25:49 2021
NAMESPACE: monitor-efk
STATUS: deployed
REVISION: 1
NOTES:
Get Fluentd build information by running these commands:

export POD_NAME=$(kubectl get pods --namespace monitor-efk -l "app.kubernetes.io/name=fluentd,app.kubernetes.io/instance=fluentd-1617186347" -o jsonpath="{.items[0].metadata.name}")
echo "curl http://127.0.0.1:24231/metrics for Fluentd metrics"
kubectl --namespace monitor-efk port-forward $POD_NAME 24231:24231

sansae@win10pro-worksp:/workspaces/fluent/charts$ kubectl get pod -n monitor-efk
NAME                                        READY   STATUS             RESTARTS   AGE
elasticsearch-master-0                      1/1     Running            0          16m
elasticsearch-master-1                      1/1     Running            0          16m
elasticsearch-master-2                      1/1     Running            0          16m
fluentd-1617190066-9g8n8                    1/1     Running            2          12m
fluentd-1617190066-9vnjh                    1/1     Running            3          12m
fluentd-1617190066-ccmth                    0/1     CrashLoopBackOff   7          12m
fluentd-1617190066-kfr96                    1/1     Running            2          12m
fluentd-1617190066-kxtmh                    1/1     Running            6          12m
fluentd-1617190066-n4566                    1/1     Running            6          12m
fluentd-1617190066-n77b8                    1/1     Running            4          12m
fluentd-1617190066-nxxq5                    1/1     Running            2          12m
fluentd-1617190066-v4dp9                    1/1     Running            2          12m
fluentd-1617190066-x4l6g                    1/1     Running            6          12m

03-04. Kibana

sansae@win10pro-worksp:/workspaces$ cd elastic
sansae@win10pro-worksp:/workspaces/elastic$ 
sansae@win10pro-worksp:/workspaces/elastic$ vi kibana/values.yaml

=========================================================
imageTag: "7.12.0"
=========================================================

sansae@win10pro-worksp:/workspaces/elastic$ helm install kibana/ --generate-name -n monitor-efk
NAME: kibana-1617190735
LAST DEPLOYED: Wed Mar 31 20:38:57 2021
NAMESPACE: monitor-efk
STATUS: deployed
REVISION: 1
TEST SUITE: None

sansae@win10pro-worksp:/workspaces/elastic$ kubectl get pod -n monitor-efk
NAME                                        READY   STATUS    RESTARTS   AGE
elasticsearch-master-0                      1/1     Running   0          15h
elasticsearch-master-1                      1/1     Running   0          15h
elasticsearch-master-2                      1/1     Running   0          15h
fluentd-1617190066-9g8n8                    1/1     Running   2          14h
fluentd-1617190066-9vnjh                    1/1     Running   3          14h
fluentd-1617190066-ccmth                    1/1     Running   8          14h
fluentd-1617190066-kfr96                    1/1     Running   2          14h
fluentd-1617190066-kxtmh                    1/1     Running   6          14h
fluentd-1617190066-n4566                    1/1     Running   6          14h
fluentd-1617190066-n77b8                    1/1     Running   4          14h
fluentd-1617190066-nxxq5                    1/1     Running   2          14h
fluentd-1617190066-v4dp9                    1/1     Running   2          14h
fluentd-1617190066-x4l6g                    1/1     Running   6          14h
kibana-1617190735-kibana-77f8f5b8c8-zxt8b   1/1     Running   0          14h

04. Dashboard 확인

sansae@win10pro-worksp:/workspaces/elastic$ kubectl get svc -n monitor-efk
NAME                            TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)             AGE
elasticsearch-master            ClusterIP   10.0.186.68    <none>        9200/TCP,9300/TCP   18m
elasticsearch-master-headless   ClusterIP   None           <none>        9200/TCP,9300/TCP   18m
fluentd-1617190066              ClusterIP   10.0.173.141   <none>        24231/TCP           14m
kibana-1617190735-kibana        ClusterIP   10.0.17.173    <none>        5601/TCP            3m4s
sansae@win10pro-worksp:/workspaces/elastic$ kubectl port-forward svc/kibana-1617190735-kibana 5601 -n monitor-efk
Forwarding from 127.0.0.1:5601 -> 5601
Forwarding from [::1]:5601 -> 5601


  • No labels
Write a comment…