「云原生」Elasticsearch + Kibana on k8s 讲解与实战操作

作者：大数据老司机 2022-10-14 07:42:50

云计算

云原生 Elasticsearch是一个基于Lucene的搜索引擎。它提供了具有HTTP Web界面和无架构JSON文档的分布式，多租户能力的全文搜索引擎。

一、概述

Elasticsearch是一个基于Lucene的搜索引擎。它提供了具有HTTP Web界面和无架构JSON文档的分布式，多租户能力的全文搜索引擎。Elasticsearch是用Java开发的，根据Apache许可条款作为开源发布。

官方文档：https://www.elastic.co/guide/en/elasticsearch/reference/master/getting-started.htmlGitHub: https://github.com/elastic/elasticsearch

二、Elasticsearch 编排部署

地址：https://artifacthub.io/packages/helm/elastic/elasticsearch

1）添加源并下载编排部署包

helm repo add elastic https://helm.elastic.co
helm pull elastic/elasticsearch --version 7.17.3
tar -xf elasticsearch-7.17.3.tgz

2）构建镜像

Elasticsearch 各版本下载地址：https://www.elastic.co/cn/downloads/past-releases#elasticsearch这里就不重新构建镜像了，这里是将远程的镜像推送到我们本地harbor，加速拉取镜像。

docker pull docker.elastic.co/elasticsearch/elasticsearch:7.17.3

docker tag docker.elastic.co/elasticsearch/elasticsearch:7.17.3  myharbor.com/bigdata/elasticsearch:7.17.3

# 上传镜像
docker push myharbor.com/bigdata/elasticsearch:7.17.3

# 删除镜像
docker rmi myharbor.com/bigdata/elasticsearch:7.17.3
crictl rmi myharbor.com/bigdata/elasticsearch:7.17.3

3）修改yaml编排

• elasticsearch/values.yaml

image:"myharbor.com/bigdata/elasticsearch"

...

...
### 去掉这几行
volumeClaimTemplate:
  accessModes:["ReadWriteOnce"]
  resources:
    requests:
      storage:30Gi
....

persistence:
  enabled:true
  labels:
    # Add default labels for the volumeClaimTemplate of the StatefulSet
    enabled:false
  annotations:{}
  accessModes:
- ReadWriteOnce
  size:1Gi
  storageClass:"elasticsearch-local-storage"
  local:
- name: elasticsearch-0
    host:"local-168-182-110"
    path:"/opt/bigdata/servers/elasticsearch/data/data1"
- name: elasticsearch-1
    host:"local-168-182-111"
    path:"/opt/bigdata/servers/elasticsearch/data/data1"
- name: elasticsearch-2
    host:"local-168-182-112"
    path:"/opt/bigdata/servers/elasticsearch/data/data1"

...

protocol: http
httpPort:9200
transportPort:9300
service:
  enabled:true
  type: NodePort
  nodePort:30920
  httpPortName:

• elasticsearch/templates/storage-class.yaml

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name:{{.Values.persistence.storageClass}}
provisioner: kubernetes.io/no-provisioner

• elasticsearch/templates/pv.yaml

{{- range .Values.persistence.local }}
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: {{ .name }}
  labels:
    name: {{ .name }}
spec:
  storageClassName: {{ $.Values.persistence.storageClass }}
  capacity:
    storage: {{ $.Values.persistence.size }}
  accessModes:
  {{- range $.Values.persistence.accessModes }}
    - {{ . | quote }}
  {{- end }}
  local:
    path: {{ .path }}
  nodeAffinity:
    required:
      nodeSelectorTerms:
        - matchExpressions:
            - key: kubernetes.io/hostname
              operator: In
              values:
                - {{ .host }}
---
{{- end }}

• elasticsearch/templates/statefulset.yaml

spec:
  volumeClaimTemplates:
    spec:
# 去掉这行
{{ toYaml .Values.volumeClaimTemplate | indent 6 }}

# 新增以下内容：
      accessModes:
      {{- range .Values.persistence.accessModes }}
      - {{ . | quote }}
      {{- end }}
      resources:
        requests:
          storage: {{ .Values.persistence.size | quote }}
    {{- if .Values.persistence.storageClass }}
    {{- if (eq "-" .Values.persistence.storageClass) }}
      storageClassName: ""
    {{- else }}
      storageClassName: "{{ .Values.persistence.storageClass }}"
    {{- end }}
    {{- end }}

4）开始部署

# 先创建本地存储目录
mkdir -p /opt/bigdata/servers/elasticsearch/data/data1
chmod -R 777/opt/bigdata/servers/elasticsearch/data/data1

helm install my-elasticsearch ./elasticsearch -n elasticsearch --create-namespace
# 查看
helm get notes my-elasticsearch -n elasticsearch
kubectl get pods,svc -n elasticsearch -owide

NOTES

NAME: my-elasticsearch
LAST DEPLOYED: Wed Oct 1223:47:172022
NAMESPACE: elasticsearch
STATUS: deployed
REVISION:1
NOTES:
1. Watch all cluster members come up.
  $ kubectl get pods --namespace=elasticsearch -l app=elasticsearch-master -w2. Test cluster health using Helm test.
  $ helm --namespace=elasticsearch test my-elasticsearch

5）测试验证

://192.168.182.110:30920/
curl http://192.168.182.110:30920/_cat/nodes
curl http://192.168.182.110:30920/_cat/health?pretty

6）elasticsearch-head

elasticsearch-head GitHub下载地址：https://github.com/mobz/elasticsearch-headGoogle 浏览器 elasticsearch-head插件：

链接：https://pan.baidu.com/s/1kYcTjBDPmSWVzsku2hEW7w?pwd=67v4提取码：67v4

7）卸载

helm uninstall my-elasticsearch -n elasticsearch
kubectl delete ns elasticsearch --force

rm -fr /opt/bigdata/servers/elasticsearch/data/data1/*
ssh local-168-182-111 "rm -fr /opt/bigdata/servers/elasticsearch/data/data1/*"
ssh local-168-182-112 "rm -fr /opt/bigdata/servers/elasticsearch/data/data1/*"

docker rmi myharbor.com/bigdata/elasticsearch:7.17.3
crictl rmi myharbor.com/bigdata/elasticsearch:7.17.3
ssh local-168-182-111 "crictl rmi myharbor.com/bigdata/elasticsearch:7.17.3"
ssh local-168-182-112 "crictl rmi myharbor.com/bigdata/elasticsearch:7.17.3"

三、Kibana 编排部署

地址：https://artifacthub.io/packages/helm/bitnami/kibana?modal=install

1）添加源并下载编排部署包

helm repo add bitnami https://charts.bitnami.com/bitnami
helm pull bitnami/kibana --version 10.2.6
tar -xf kibana-10.2.6.tgz

2）构建镜像

这里也不重新构建镜像了，只是将镜像推送到本地harbor加速，【注意】版本号需要跟ES相对应，版本不一样，可能会存储不兼容的情况。

docker pull docker.io/bitnami/kibana:7.17.3
docker tag docker.io/bitnami/kibana:7.17.3 myharbor.com/bigdata/kibana:7.17.3

# 上传镜像
docker push myharbor.com/bigdata/kibana:7.17.3

# 删除镜像
docker rmi myharbor.com/bigdata/kibana:7.17.3
crictl rmi myharbor.com/bigdata/kibana:7.17.3

3）修改yaml编排

• kibana/values.yaml

image:
  registry: myharbor.com
  repository: bigdata/kibana
  tag: 8.4.3-debian-11-r1

...

replicaCount: 1

...

persistence:
  enabled: true
  accessModes:
    - ReadWriteOnce
  size: 1Gi
  storageClass: "kibana-local-storage"
  local:
  - name: kibana-0
    host: "local-168-182-111"
    path: "/opt/bigdata/servers/kibana/data/data1"

...

service:
  ports:
    http: 5601
  type: NodePort
  nodePorts:
    http: "30601"

...

elasticsearch:
  hosts:
    - elasticsearch-master.elasticsearch
  port: "9200"

• kibana/templates/values.yaml

{{- range .Values.persistence.local }}
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: {{ .name }}
  labels:
    name: {{ .name }}
spec:
  storageClassName: {{ $.Values.persistence.storageClass }}
  capacity:
    storage: {{ $.Values.persistence.size }}
  accessModes:
  {{- range $.Values.persistence.accessModes }}
    - {{ . | quote }}
  {{- end }}
  local:
    path: {{ .path }}
  nodeAffinity:
    required:
      nodeSelectorTerms:
        - matchExpressions:
            - key: kubernetes.io/hostname
              operator: In
              values:
                - {{ .host }}
---
{{- end }}

• kibana/templates/storage-class.yaml

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: {{ .Values.persistence.storageClass }}
provisioner: kubernetes.io/no-provisioner

• kibana/templates/pv.yaml

{{- range .Values.persistence.local }}
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: {{ .name }}
  labels:
    name: {{ .name }}
spec:
  storageClassName: {{ $.Values.persistence.storageClass }}
  capacity:
    storage: {{ $.Values.persistence.size }}
  accessModes:
  {{- range $.Values.persistence.accessModes }}
    - {{ . | quote }}
  {{- end }}
  local:
    path: {{ .path }}
  nodeAffinity:
    required:
      nodeSelectorTerms:
        - matchExpressions:
            - key: kubernetes.io/hostname
              operator: In
              values:
                - {{ .host }}
---
{{- end }}

4）开始部署

# 先创建本地存储目录
mkdir -p /opt/bigdata/servers/kibana/data/data1
chmod -R 777 /opt/bigdata/servers/kibana/data/data1

helm install my-kibana ./kibana -n kibana --create-namespace
# 查看
helm get notes my-kibana -n kibana 
kubectl get pods,svc -n kibana -owide

NOTES

NAME: my-kibana
LAST DEPLOYED: Thu Oct 13 22:43:30 2022
NAMESPACE: kibana
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
CHART NAME: kibana
CHART VERSION: 10.2.6
APP VERSION: 8.4.3

** Please be patient while the chart is being deployed **######################################################################################################
### ERROR: You did not provide the Elasticsearch external host or port in your 'helm install' call ###
######################################################################################################

Complete your Kibana deployment by running:

  helm upgrade --namespace kibana my-kibana my-repo/kibana \
    --set elasticsearch.hosts[0]=YOUR_ES_HOST,elasticsearch.port=YOUR_ES_PORT

Replacing "YOUR_ES_HOST" and "YOUR_ES_PORT" placeholders by the proper values of your Elasticsearch deployment.

5）测试验证

web：http://192.168.182.111:30601

6）卸载

helm uninstall my-kibana -n kibana
kubectl delete ns kibana --force

ssh local-168-182-111 rm -fr /opt/bigdata/servers/kibana/data/data1/*

docker rmi myharbor.com/bigdata/bigdata:8.4.3-debian-11-r1
crictl rmi myharbor.com/bigdata/bigdata:8.4.3-debian-11-r1

elasticsearch-on-k8s

下载地址：https://gitee.com/hadoop-bigdata/elasticsearch-on-k8skibana-on-k8s

下载地址：https://gitee.com/hadoop-bigdata/kibana-on-k8s