ElasticSearch深度集成kubernetes

Posted by Zeusro on December 8, 2018

安装和配置

自制带插件的ES镜像

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
FROM elasticsearch:6.5.0
#或者手动下载后然后安装也行
# COPY elasticsearch-analysis-ik-6.5.0.zip /
# elasticsearch-plugin install --batch file:///elasticsearch-analysis-ik-6.5.0.zip
    #IK Analyzer是一个开源的,基于java语言开发的中文分词工具包。是开源社区中处理中分分词非常热门的插件。 
RUN elasticsearch-plugin install --batch https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v6.5.0/elasticsearch-analysis-ik-6.5.0.zip && \
    # 拼音分词器
    elasticsearch-plugin install --batch https://github.com/medcl/elasticsearch-analysis-pinyin/releases/download/v6.5.0/elasticsearch-analysis-pinyin-6.5.0.zip && \    
    # Smart Chinese Analysis Plugin
    elasticsearch-plugin install analysis-icu && \
    # 日文分词器
    elasticsearch-plugin install analysis-kuromoji && \
    # 语音分析
    elasticsearch-plugin install analysis-phonetic && \
    # 计算字符哈希
    elasticsearch-plugin install  mapper-murmur3 && \
    # 在_source中提供size字段
    elasticsearch-plugin install mapper-size

编排文件

  • RBAC 相关内容
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
apiVersion: v1
kind: ServiceAccount
metadata:
  name: elasticsearch-admin
  namespace: default
  labels:
    app: elasticsearch
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  namespace: default
  name: elasticsearch
  labels:
    app: elasticsearch
subjects:
- kind: ServiceAccount
  name: elasticsearch-admin
  namespace: default
  apiGroup: ""
roleRef:
  kind: ClusterRole
  name: elasticsearch
  apiGroup: ""
  • 主体程序

存储使用了hostpath,需要先在宿主机闯将目录,并赋予适当的权限,不然会出错

1
2
3
mkdir -p /root/kubernetes/$(namespace)/elasticsearch/data
sudo chmod 775  /root/kubernetes/$(namespace)/elasticsearch/data -R
chown 1000:0  /root/kubernetes/$(namespace)/elasticsearch/data -R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
# kubectl get po -l app=myelasticsearch -o wide  -n default
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: myelasticsearch
  namespace: default
  labels:
    app: myelasticsearch    
    elasticsearch-role: all
spec:
  updateStrategy:
    rollingUpdate:
      partition: 0
    type: RollingUpdate
  serviceName: elasticsearch-master
  replicas: 2
  selector:
    matchLabels:
      app: myelasticsearch
      elasticsearch-role: all
  template:
    metadata:
      labels:
        app: myelasticsearch                     
        elasticsearch-role: all 
    spec:
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: elasticsearch-test-ready
                operator: Exists
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:              
              - key: 'app'
                operator: In
                values:
                - myelasticsearch
              - key: 'elasticsearch-role'
                operator: In
                values:
                - all                 
            topologyKey: "kubernetes.io/hostname"
            namespaces: 
            - default    
      serviceAccountName: elasticsearch-admin    
      terminationGracePeriodSeconds: 180
      # Elasticsearch requires vm.max_map_count to be at least 262144.
      # If your OS already sets up this number to a higher value, feel free
      # to remove this init container.
      initContainers:
      - image: alpine:3.6
        command: ["/sbin/sysctl", "-w", "vm.max_map_count=262144"]
        name: elasticsearch-init
        securityContext:
          privileged: true      
      imagePullSecrets:
        - name: vpc-shenzhen      
      containers:
      - image: elasticsearch:6.5.0-plugin-in-remote-ik
        name: elasticsearch
        resources:
          # need more cpu upon initialization, therefore burstable class
          limits:
            # cpu: 2
            memory: 4Gi
          requests:
            # cpu: 1
            memory: 1Gi
        ports:
        - name: restful
          containerPort: 9200          
          protocol: TCP
        - name: discovery
          containerPort: 9300          
          protocol: TCP
        readinessProbe:
          failureThreshold: 3
          initialDelaySeconds: 2
          periodSeconds: 10
          successThreshold: 1
          tcpSocket:
            port: 9200
          timeoutSeconds: 1     
        # livenessProbe:
        #   failureThreshold: 3
        #   initialDelaySeconds: 7
        #   periodSeconds: 10
        #   successThreshold: 1
        #   tcpSocket:
        #     port: 9200
        #   timeoutSeconds: 1           
        volumeMounts:
        - name: host
          mountPath: /usr/share/elasticsearch/data
        env:
        - name: "NAMESPACE"
          valueFrom:
            fieldRef:
              fieldPath: metadata.namespace
        # - name: "ES_JAVA_OPTS"
        #   value: "-Xms256m -Xmx256m"              
        - name: "cluster.name"
          value: "myelasticsearch"
        - name: "bootstrap.memory_lock"
          value: "true"
        - name: "discovery.zen.ping.unicast.hosts"
          value: "myelasticsearch"
        - name: "discovery.zen.minimum_master_nodes"
          value: "2"
        - name: "discovery.zen.ping_timeout"
          value: "5s"
          #因为是测试,所以master,data,ingest都混用          
        - name: "node.master"
          value: "true"
        - name: "node.data"
          value: "true"
        - name: "node.ingest"
          value: "true"
        - name: xpack.monitoring.collection.enabled
          value: "true" 
        - name: node.name
          valueFrom: 
            fieldRef:
              fieldPath: metadata.name 
              # fieldPath: spec.nodeName 
        securityContext: 
          privileged: true        
      volumes:
      - name: host
        hostPath:
          path: /root/kubernetes/default/myelasticsearch/data
          type: DirectoryOrCreate
---
kind: Service
apiVersion: v1
metadata:
  labels:
    app: myelasticsearch
    elasticsearch-role: all
  name: myelasticsearch
  namespace: default
spec:
  ports:
    - name: discovery
      port: 9300
      targetPort: discovery   
    - name: restful
      port: 9200
      protocol: TCP
      targetPort: restful      
  selector:
    app: myelasticsearch
    elasticsearch-role: all
  type: NodePort

这个编排精髓的一点在于用了节点affinity使每一个节点最多会运行一个容器,确保了高可用.

如果要吧节点的角色再抽取出来,那么其实抽取一个service作为相互发现的,即可.

1
2
3
4
5
6
7
8
9
10
11
12
13
kind: Service
apiVersion: v1
metadata:
  labels:
    app: elasticsearch
  name: elasticsearch-discovery
  namespace: default
spec:
  ports:
    - port: 9300
      targetPort: discovery
  selector:
    app: myelasticsearch
  1. 在Kubernetes上部署Elasticsearch集群
  2. 重要配置的修改
  3. Install Elasticsearch with Docker
  4. 学习Elasticsearch之4:配置一个3节点Elasticsearch集群(不区分主节点和数据节点)
  5. 谈一谈Elasticsearch的集群部署
  6. 官方docker镜像构建项目
  7. Elasticsearch模块功能之-自动发现(Discovery)
  8. 订阅费用
  9. 故障转移
  10. 节点配置
  11. Elasticsearch 5.X集群多节点角色配置深入详解
  12. elasticsearch-cloud-kubernetes
  13. [干货 吃透Elasticsearch 堆内存](https://blog.csdn.net/laoyang360/article/details/79998974)

[x] 配置账户

  • 插件安装

elasticsearch-docker-plugin-management

安装插件

GET /_cat/plugins?v&s=component&h=name,component,version,description

压力测试

1
2
3
esrally configure
esrally list tracks
esrally --pipeline=benchmark-only --target-hosts=127.0.0.1:9200 --track=geonames

datastore.type = elasticsearch datastore.host = 127.0.0.1 datastore.port = 9200 datastore.secure = False

通过 Elasticsearch 官方提供的 benchmark 脚本 rally

ElasticSearch集群的维护

更新的时候务必使用灰度更新,从序号最大的镜像开始更新,不然分片丢失了,相信我,你会死的很惨

1
2
3
kubectl patch statefulset elasticsearch -p '{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":0}}}}' -n test
kubectl patch statefulset elasticsearch -p '{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":1}}}}' -n test
kubectl patch statefulset elasticsearch -p '{"spec":{"updateStrategy":{"type":"RollingUpdate","rollingUpdate":{"partition":2}}}}' -n test