【k8s】【ELK】【zookeeper+kafka+efak】日志环境部署

1、日志收集基本概念

k8s中pod的路径：
containers log: /var/log/containers/*.log
Pod log： /var/log/pods
docker log: /var/lib/docker/containers/*/*.log

如何收集日志

使用 EFK+Logstash+Kafka
在这里插入图片描述

1、filebeat读取容器中的日志，然后写入Kafka集群中对应的topic（重点在于filebeat如何收集容器日志）；
2、logstash从Kafka中读取数据，然后对数据进行清洗，而后写入ES对应的索引中；
3、Kibana匹配对应的ES索引，进行日志展示，分析

收集K8S哪些日志?

在这里插入图片描述

2、交付Elastic集群

#创建ns 、 docker-registry
kubectl create namespace logging

kubectl create secret docker-registry harbor-admin \
--docker-username=admin \
--docker-password=Harbor12345 \
--docker-server=harbor.oldxu.net \
-n logging

2.1 ES集群的构建

传统方式部署ES集群，参考：
https://blog.csdn.net/weixin_42890981/article/details/126692731 
和
https://blog.csdn.net/opensystem123/article/details/128131030

1、 ES 集群是由多个节点组成的，通过 cluster.name设置ES集群名称，同时用于区分其它的ES集群。
2、每个节点通过 node.name 参数来设定所在集群的节点名称。
3、节点使用 discovery.send_hosts 参数来设定集群节点的列表。

4、集群在第一次启动时，需要初始化，同时需要指定参与选举的master节点IP，或节点名称。
5、每个节点可以通过 node.master:true 设定为master角色，通过 node.data:true 设定为data角色。

demo：
在这里插入图片描述

2.2 交付ES-Service

#docker pull docker.elastic.co/elasticsearch/elasticsearch:7.17.6
docker pull elasticsearch:7.17.6
docker tag 5fad10241ffd harbor.oldxu.net/base/elasticsearch:7.17.6
docker push harbor.oldxu.net/base/elasticsearch:7.17.6

01-es-svc.yaml

#创建es-headlessService，为每个ES Pod设定固定的DNS名称，无论它是Master或是Data，易或是Coordinating
apiVersion: v1
kind: Service
metadata:
  name: es-svc
  namespace: logging
spec:
  selector:
    app: es
  clusterIP: None
  ports:
  - name: cluster
    port: 9200
    targetPort: 9200
  - name: transport
    port: 9300
    targetPort: 9300

2.3 交付ES-Master节点

1、ES无法使用root直接启动，需要授权数据目录UID=1000，同时还需要持久化/usr/share/elasticsearch/data
2、ES所有节点都需要设定 vm.max_map_count 内核参数以及ulimit；

3、ES启动是通过ENV环境变量传参来完成的
3.1、集群名称、节点名称、角色类型
3.2、 discovery.seed_hosts 集群地址列表
3.3、 cluster.initial_master_nodes 初始集群参与选举的master节点名称；

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: es-master
  namespace: logging
spec:
  serviceName: "es-svc"
  replicas: 3
  selector:
    matchLabels:
      app: es
      role: master
  template:
    metadata:
      labels:
        app: es
        role: master
    spec:
      imagePullSecrets:
      - name: harbor-admin
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:
              - key: app
                operator: In
                values: ["es"]
              - key: role
                operator: In
                values: ["master"]
            topologyKey: "kubernetes.io/hostname"
            
      initContainers:
      - name: fix-permissions
        image: busybox
        command: ["sh","-c","chown -R 1000:1000 /usr/share/elasticsearch/data ; sysctl -w vm.max_map_count=262144; ulimit -n 65536"]        
        securityContext:
          privileged: true
        volumeMounts:
        - name: data
          mountPath: /usr/share/elasticsearch/data
          
      containers:
      - name: es
        image: harbor.oldxu.net/base/elasticsearch:7.17.6
        resources:
          limits:
            cpu: 1000m
            memory: 1024Mi
        ports:
        - name: cluster
          containerPort: 9200
        - name: transport
          containerPort: 9300
        
        volumeMounts:
        - name: data
          mountPath: /usr/share/elasticsearch/data
        env:
        - name: ES_JAVA_OPTS
          value: "-Xms512m -Xmx512m"        
        - name: cluster.name
          value: es-oldxu
        - name: node.name
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        - name: node.master
          value: "true"
        - name: node.data
          value: "false"
        - name: discovery.seed_hosts
          value: "es-master-0.es-svc,es-master-1.es-svc,es-master-2.es-svc"
        - name: cluster.initial_master_nodes
          value: "es-master-0,es-master-1,es-master-2"
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: ["ReadWriteOnce"]   
      storageClassName: "nfs"
      resources:
        requests:
          storage: 25Gi

2.4交付ES-Data节点

1、ES无法使用root直接启动，需要授权数据目录UID=1000，同时还需要持久化 /usr/share/elasticsearch/data
2、ES所有节点都需要设定 vm.max_map_count 内核参数以及ulimit；

3、ES启动是通过ENV环境变量传参来完成的
3.1、集群名称、节点名称、角色类型
3.2、 discovery.seed_hosts 集群节点地址，任意填写Master节点域名
#资源内存有限，部署2个ES-data节点

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: es-data
  namespace: logging
spec:
  serviceName: "es-svc"
  replicas: 2
  selector:
    matchLabels:
      app: es
      role: data
  template:
    metadata:
      labels:
        app: es
        role: data
    spec:
      imagePullSecrets:
      - name: harbor-admin
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:
              - key: app
                operator: In
                values: ["es"]
              - key: role
                operator: In
                values: ["data"]
            topologyKey: "kubernetes.io/hostname"
      initContainers:
      - name: fix-permissions
        image: busybox
        command: ["sh","-c","chown -R 1000:1000 /usr/share/elasticsearch/data ; sysctl -w vm.max_map_count=262144; ulimit -n 65536"]
        securityContext:
          privileged: true
        volumeMounts:
        - name: data
          mountPath: /usr/share/elasticsearch/data
      
      containers:
      - name: es
        image: harbor.oldxu.net/base/elasticsearch:7.17.6
        resources:
          limits:
            cpu: 1000m
            memory: 1024Mi
        ports:
        - name: cluster
          containerPort: 9200
        - name: transport
          containerPort: 9300
        volumeMounts:
        - name: data
          mountPath: /usr/share/elasticsearch/data
        env:
        - name: ES_JAVA_OPTS
          value: "-Xms512m -Xmx512m"
        - name: cluster.name
          value: es-oldxu
        - name: node.name
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        - name: node.master
          value: "false"
        - name: node.data
          value: "true"
        - name: discovery.seed_hosts
          es-master-0.es-svc,es-master-1.es-svc,es-master-2.es-svc
          #value: "es-master-0,es-master-1,es-master-2"  

  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: ["ReadWriteOnce"]
      storageClassName: "nfs"
      resources:
        requests:
          storage: 20Gi

2.5 验证ES集群

[root@master01 02-es-data-service]# dig @10.96.0.10 es-svc.logging.svc.cluster.local +short
10.244.2.169
10.244.0.3
10.244.1.213
10.244.1.212
10.244.2.164

在这里插入图片描述

通过curl访问ES，检查ES集群是否正常（如果仅交付Master，没有data节点，集群状态可能会Red，因为没有数据节点进行数据存储；）

curl -XGET "http://10.244.1.213:9200/_cluster/health?pretty"
curl -XGET "http://10.244.1.213:9200/_cat/nodes?/v"
curl -XGET "http://10.244.1.213:9200/_cat/shards?v"

在这里插入图片描述

注意： yaml文件的- name: discovery.seed_hosts 的值，否则也会是red。

在这里插入图片描述

3.交付Kibana可视化

3.1 交付Kibana（dp、svc、ingress）

docker pull kibana:7.17.6
docker tag kibana:7.17.6 harbor.oldxu.net/base/kibana:7.17.6
docker push harbor.oldxu.net/base/kibana:7.17.6

#说明：
#cat /etc/kibana/kibana.yml
server.port: 5601           #kibana默认监听端口
server.host: "0.0.0.0"      #kibana监听地址段
elasticsearch.hosts: ["http://localhost:9200"]
# kibana丛coordinating节点获取数据
i18n.locale: "zh-CN"        #kibana汉化

01-kibana-dp.yam

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kibana
  namespace: logging
spec:
  replicas: 1
  selector:
    matchLabels:
      app: kibana
  template:
    metadata:
      labels:
        app: kibana
    spec:
      imagePullSecrets:
      - name: harbor-admin
      containers:
      - name: kibana
        image: harbor.oldxu.net/base/kibana:7.17.6
        resources:
          limits:
            cpu: 1000m
        ports:
        - containerPort: 5601
        env:
        - name: ELASTICSEARCH_HOSTS
          value: '["http:/es-data-0.es-svc:9200","http:/es-data-1.es-svc:9200"]'        
        - name: I18N_LOCALE
          value: "zh-CN"
        - name: SERVER_PUBLICBASEURL
          value: "http://kibana.oldxu.net"

02-kibana-svc.yam

apiVersion: v1
kind: Service
metadata:
  name: kibana-svc
  namespace: logging
spec:
  selector:
    app: kibana
  ports:
  - name: web
    port: 5601
    targetPort: 5601

03-kibana-ingress.yam

#apiVersion: networking.k8s.io/v1
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
  name: kibana-ingress
  namespace: logging
spec:
  ingressClassName: "nginx"
  rules:
  - host: "kibana.oldxu.net"
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          serviceName: kibana-svc
          servicePort: 5601
          #service:
          #  name: kibana-svc
          #  port:
          #    number: 5601

3.2 访问kibana

在这里插入图片描述

4、交付Zookeeper集群至K8S

镜像： harbor.oldxu.net/base/zookeeper:3.8.0
变量 ZOOK_SERVERS ：zk集群节点通信地址

01-zk-svc.yaml

apiVersion: v1
kind: Service
metadata:
  name: zk-svc
  namespace: logging
spec:
  clusterIP: None
  selector: 
    app: zk
  ports:
  - name: client
    port: 2181
    targetPort: 2181
  - name: leader-follwer
    port: 2888
    targetPort: 2888
  - name: selection
    port: 3888
    targetPort: 3888

02-zk-sts.yaml

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: zookeeper
  namespace: logging
spec:
  serviceName: "zk-svc"
  replicas: 3
  selector:
    matchLabels:
      app: zk
  template:
    metadata:
      labels:
        app: zk
    spec:
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            - labelSelector:
                matchExpressions:
                - key: app
                  operator: In
                  values: ["zk"]
              topologyKey: "kubernetes.io/hostname" 
      imagePullSecrets:
      - name: harbor-admin
      containers:
      - name: zk
        image: harbor.oldxu.net/base/zookeeper:3.8.0
        imagePullPolicy: IfNotPresent
        ports:
        - name: client
          containerPort: 2181
        - name: leader-follwer
          containerPort: 2888
        - name: selection
          containerPort: 3888
        env:
        - name: ZOOK_SERVERS                
          value: "server.1=zookeeper-0.zk-svc.logging.svc.cluster.local:2888:3888 server.2=zookeeper-1.zk-svc.logging.svc.cluster.local:2888:3888 server.3=zookeeper-2.zk-svc.logging.svc.cluster.local:2888:3888"
        readinessProbe:
          exec:
            command:
            - "/bin/bash"
            - "-c"
            - '[[ "$(/zookeeper/bin/zkServer.sh status 2>/dev/null|grep 2181)" ]] && exit 0 || exit 1'
          initialDelaySeconds: 5
        livenessProbe:
          exec:
            command:
            - "/bin/bash"
            - "-c"
            - '[[ "$(/zookeeper/bin/zkServer.sh status 2>/dev/null|grep 2181)" ]] && exit 0 || exit 1'
        volumeMounts:
        - name: data
          mountPath: /data
          
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: ["ReadWriteMany"]
      storageClassName: "nfs" 
      resources:
        requests:
          storage: 21Gi

验证zk集群

kubectl exec -it -n logging zookeeper-0 -- /zookeeper/bin/zkServer.sh status
kubectl exec -it -n logging zookeeper-1 -- /zookeeper/bin/zkServer.sh status
kubectl exec -it -n logging zookeeper-2 -- /zookeeper/bin/zkServer.sh status

在这里插入图片描述

5、交付Kafka集群至K8S

镜像名称及tag： harbor.oldxu.net/base/kafka:2.12
变量： ZOOK_SERVERS ：Kafka元数据存储ZK，需明确指定Zookeeper集群地址

01-kafka-svc.yaml

apiVersion: v1
kind: Service
metadata:
  name: kafka-svc
  namespace: logging
spec:
  clusterIP: None
  selector:
    app: kafka
  ports:
  - name: client
    port: 9092
    targetPort: 9092
  - name: jmx
    port: 9999
    targetPort: 9999

02-kafka-sts.yaml

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: kafka
  namespace: logging
spec:
  serviceName: "kafka-svc"
  replicas: 3
  selector:
    matchLabels:
      app: kafka
  template:
    metadata:
      labels:
        app: kafka
    spec:
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            - labelSelector:
                matchExpressions:
                  - key: app
                    operator: In
                    value: ["kafka"]
              topologyKey: "kubernetes.io/hostname"
      imagePullSecrets:
      - name: harbor-admin
      containers:
      - name: kafka
        image: harbor.oldxu.net/base/kafka:2.12
        imagePullPolicy: IfNotPresent
        ports:
        - name: client
          containerPort: 9092
        - name: jmxport
          containerPort: 9999
        env:
        - name: ZOOK_SERVERS
          value: "zookeeper-0.zk-svc:2181,zookeeper-1.zk-svc:2181,zookeeper-2.zk-svc:2181"        
        readinessProbe:
          tcpSocket:
            port: 9092
          initialDelaySeconds: 5
        livenessProbe:
          tcpSocket:
            port: 9092
          initialDelaySeconds: 5
        volumeMounts:
        - name: data
          mountPath: /data    

  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: ["ReadWriteMany"]
      storageClassName: "nfs"
      resources:
        requests:
          storage: 20Gi

验证Kafka集群

#1、登陆任意kafka Pod，然后创建topic主题
[root@master01 ~]# kubectl exec -it -n logging kafka-0 -- bash

root@kafka-0:/# /kafka/bin/kafka-topics.sh --create --zookeeper zookeeper-0.zk-svc:2181,zookeeper-1.zk-svc:2181,zookeeper-2.zk-svc:2181 --partitions 1 --replication-factor 3 --topic oldl33
Created topic oldl33.

#2、消息发布
[root@master01 ~]# kubectl exec -it -n logging kafka-1 -- bash
root@kafka-1:/# /kafka/bin/kafka-console-producer.sh \
> --broker-list kafka-0.kafka-svc:9092,kafka-1.kafka-svc:9092,kafka-2.kafka-svc:9092 \
> --topic oldl33
>
>deploy kafka ok
>heloo kafka on k8s
>

#3、消息订阅（可以登陆其他Pod进行消息订阅）
[root@master01 ~]# kubectl exec -it -n logging kafka-2 -- bash
root@kafka-2:/# ls
bin  boot  data  dev  entrypoint.sh  etc  home	kafka  lib  lib64  media  mnt  opt  proc  root	run  sbin  srv	sys  tmp  usr  var
root@kafka-2:/# 
root@kafka-2:/# /kafka/bin/kafka-console-consumer.sh \
> --bootstrap-server kafka-0.kafka-svc:9092,kafka-1.kafka-svc:9092,kafka-2.kafka-svc:9092 \
> --topic oldl33 \
> --from-beginning

deploy kafka ok
heloo kafka on k8s

6、交付EFAK至K8S

镜像名称及tag： harbor.oldxu.net/base/efak:3.0
变量： ZOOK_SERVERS ：efak需要连接Zk获取zk和kafka集群各种状态指标信息

01-efak-dp.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: efak
  namespace: logging
spec:
  replicas: 1
  selector:
    matchLabels:
      app: efak
  template:
    metadata:
      labels:
        app: efak
    spec:
      imagePullSecrets:
      - name: harbor-admin
      containers:
      - name: efak
        image: harbor.oldxu.net/base/efak:3.0
        imagePullPolicy: IfNotPresent
        
        ports:
        - name: http
          containerPort: 8048
        env:
        - name: ZOOK_SERVERS
          value: "zookeeper-0.zk-svc:2181,zookeeper-1.zk-svc:2181,zookeeper-2.zk-svc:2181"

02-efak-svc.yaml

apiVersion: v1
kind: Service
metadata:
  name: efak-svc
  namespace: logging
spec:
  selector:
    app: efak
  ports:
  - port: 8048
    targetPort: 8048

03-efak-ingress.yaml

apiVersion: networking.k8s.io/betav1
kind: Ingress
metadata:
  name: efak-ingress
  namespace: logging
spec:
  ingressClassName: "nginx"
  rules:
  - host: "efak.oldxu.net"
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          serviceName: efak-svc
          servicePort: 8048
          #service:
          #  name: efak-svc
          #  port:
          #    number: 8048

访问eagle

admin / 123456
zk和kafka状态：
在这里插入图片描述

END

备注：

以前的博文有详细部署zookeeper 、kafka、efak的描述
https://blog.csdn.net/Nightwish5/article/details/130389581 《zookeeper》
https://blog.csdn.net/Nightwish5/article/details/130402955 《kafka集群，efak 》