一.kafka介紹

kafka是一個分散式、多副本、多訂閱者、分割槽的，基於zoopkeeper協調的分散式日誌系統。其主要特點為：
1.以時間複雜度為O(1)的方式提供訊息持久化能力，即使對TB級以上的資料也能保證常數時間的訪問效能。
2.高吞吐率。即使在非常廉價的商用機器上也能做到單機支援每秒100K訊息的傳輸。
3.支援kafka server間的訊息分割槽以及分散式消費，同時保證每個partition內的訊息順序和傳輸。
4.同時支援離線資料處理和實時資料處理。

二.應用場景

1.日誌收集
2.資料推送
3.作為大緩衝區使用
4.服務中介軟體

三.應用架構

如上圖所示，一個kafka叢集包含若干個Producer(伺服器日誌、業務資料、Web前端產生的page view等)，若干個Broker(kafka支援水平擴充套件，一般broker數量越多，叢集的吞吐量越大)，若干個consumer group，一個zookeeper叢集(kafka通過zookeeper管理叢集配置、選舉leader、consumer group等發生變化時進行rebalance)。

3.1 名詞解釋

• broker
  訊息中介軟體處理節點(伺服器)，一個節點就是一個broker，一個kafka叢集由一個或多個broker組成

• Topic
  kafka對訊息進行歸類，傳送到叢集的每一條訊息都要指定一個topic

• Partition
  物理上的概念，每個topic包含一個或多個partition，一個partition對應一個資料夾，這個資料夾下儲存partition的資料和索引檔案，每個partition內部是有序的。

• Producer
  生產者，負責釋出訊息到broker

• Consumer
  消費者，從broker讀取訊息

• ConsumerGroup
  每個consumer屬於一個特定的consumer group，可為每個consumer指定group name，若不指定，則屬於預設的group，一條訊息可以傳送到不同的consumer group，但一個consumer group中只能有一個consumer能消費這條訊息。

四.kubernetes叢集部署kafka

4.1 部署前準備

• 建立好的至少三個節點的kubernetes叢集(這裡我們使用的版本1.13.10)
• 建立好的儲存類StorageClass(這裡我們使用的是cephfs)

4.2 部署yaml檔案

1.部署zookeeper的yaml檔案

[root@k8s001 kafka]# cat zookeeper.yaml 
apiVersion: v1
kind: Service
metadata:
  name: zk-hs
  namespace: kafka
  labels:
    app: zk
spec:
  ports:
  - port: 2888
    name: server
  - port: 3888
    name: leader-election
  clusterIP: None
  selector:
    app: zk
---
apiVersion: v1
kind: Service
metadata:
  name: zk-cs
  namespace: kafka
  labels:
    app: zk
spec:
  ports:
  - port: 2181
    name: client
  selector:
    app: zk
---
apiVersion: policy/v1beta1
kind: PodDisruptionBudget
metadata:
  name: zk-pdb
  namespace: kafka
spec:
  selector:
    matchLabels:
      app: zk
  maxUnavailable: 1
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: zk
  namespace: kafka
spec:
  selector:
    matchLabels:
      app: zk
  serviceName: zk-hs
  replicas: 3
  updateStrategy:
    type: RollingUpdate
  podManagementPolicy: Parallel
  template:
    metadata:
      labels:
        app: zk
    spec:
      nodeSelector:
          travis.io/schedule-only: "kafka"
      tolerations:
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "NoSchedule"
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "NoExecute"
        tolerationSeconds: 3600
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "PreferNoSchedule"
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            - labelSelector:
                matchExpressions:
                  - key: "app"
                    operator: In
                    values:
                    - zk
              topologyKey: "kubernetes.io/hostname"
      containers:
      - name: kubernetes-zookeeper
        imagePullPolicy: IfNotPresent
        image: fastop/zookeeper:3.4.10
        resources:
          requests:
            memory: "200Mi"
            cpu: "0.1"
        ports:
        - containerPort: 2181
          name: client
        - containerPort: 2888
          name: server
        - containerPort: 3888
          name: leader-election
        command:
        - sh
        - -c
        - "start-zookeeper \
          --servers=3 \
          --data_dir=/var/lib/zookeeper/data \
          --data_log_dir=/var/lib/zookeeper/data/log \
          --conf_dir=/opt/zookeeper/conf \
          --client_port=2181 \
          --election_port=3888 \
          --server_port=2888 \
          --tick_time=2000 \
          --init_limit=10 \
          --sync_limit=5 \
          --heap=512M \
          --max_client_cnxns=60 \
          --snap_retain_count=3 \
          --purge_interval=12 \
          --max_session_timeout=40000 \
          --min_session_timeout=4000 \
          --log_level=INFO"
        readinessProbe:
          exec:
            command:
            - sh
            - -c
            - "zookeeper-ready 2181"
          initialDelaySeconds: 10
          timeoutSeconds: 5
        livenessProbe:
          exec:
            command:
            - sh
            - -c
            - "zookeeper-ready 2181"
          initialDelaySeconds: 10
          timeoutSeconds: 5
        volumeMounts:
        - name: datadir
          mountPath: /var/lib/zookeeper
      # 這裡我們需要將runAsuser和fsGroup使用者調整為0，也就是管理員使用者允許，否則會提示許可權的報錯
      securityContext:
        runAsUser: 0
        fsGroup: 0
  volumeClaimTemplates:
  - metadata:
      name: datadir
    spec:
      accessModes: [ "ReadWriteMany" ]
      storageClassName: cephfs
      resources:
        requests:
          storage: 20Gi
 

2.部署kafka的yaml檔案

[root@k8s001 kafka]# cat kafka.yaml 
---
apiVersion: v1
kind: Service
metadata:
  name: kafka-svc
  namespace: kafka
  labels:
    app: kafka
spec:
  ports:
  - port: 9092
    name: server
  clusterIP: None
  selector:
    app: kafka
---
apiVersion: policy/v1beta1
kind: PodDisruptionBudget
metadata:
  name: kafka-pdb
  namespace: kafka
spec:
  selector:
    matchLabels:
      app: kafka
  minAvailable: 2
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: kafka
  namespace: kafka
spec:
  selector:
     matchLabels:
        app: kafka
  serviceName: kafka-svc
  replicas: 3
  template:
    metadata:
      labels:
        app: kafka
    spec:
      nodeSelector:
          travis.io/schedule-only: "kafka"
      tolerations:
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "NoSchedule"
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "NoExecute"
        tolerationSeconds: 3600
      - key: "travis.io/schedule-only"
        operator: "Equal"
        value: "kafka"
        effect: "PreferNoSchedule"
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            - labelSelector:
                matchExpressions:
                  - key: "app"
                    operator: In
                    values: 
                    - kafka
              topologyKey: "kubernetes.io/hostname"
        podAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
             - weight: 1
               podAffinityTerm:
                 labelSelector:
                    matchExpressions:
                      - key: "app"
                        operator: In
                        values: 
                        - zk
                 topologyKey: "kubernetes.io/hostname"
      terminationGracePeriodSeconds: 300
      containers:
      - name: k8s-kafka
        imagePullPolicy: IfNotPresent
        image: fastop/kafka:2.2.0
        resources:
          requests:
            memory: "600Mi"
            cpu: 500m
        ports:
        - containerPort: 9092
          name: server
        command:
        - sh
        - -c
        - "exec kafka-server-start.sh /opt/kafka/config/server.properties --override broker.id=${HOSTNAME##*-} \
          --override listeners=PLAINTEXT://:9092 \
          --override zookeeper.connect=zk-0.zk-hs.kafka.svc.cluster.local:2181,zk-1.zk-hs.kafka.svc.cluster.local:2181,zk-2.zk-hs.kafka.svc.cluster.local:2181 \
          --override log.dir=/var/lib/kafka \
          --override auto.create.topics.enable=true \
          --override auto.leader.rebalance.enable=true \
          --override background.threads=10 \
          --override compression.type=producer \
          --override delete.topic.enable=false \
          --override leader.imbalance.check.interval.seconds=300 \
          --override leader.imbalance.per.broker.percentage=10 \
          --override log.flush.interval.messages=9223372036854775807 \
          --override log.flush.offset.checkpoint.interval.ms=60000 \
          --override log.flush.scheduler.interval.ms=9223372036854775807 \
          --override log.retention.bytes=-1 \
          --override log.retention.hours=168 \
          --override log.roll.hours=168 \
          --override log.roll.jitter.hours=0 \
          --override log.segment.bytes=1073741824 \
          --override log.segment.delete.delay.ms=60000 \
          --override message.max.bytes=1000012 \
          --override min.insync.replicas=1 \
          --override num.io.threads=8 \
          --override num.network.threads=3 \
          --override num.recovery.threads.per.data.dir=1 \
          --override num.replica.fetchers=1 \
          --override offset.metadata.max.bytes=4096 \
          --override offsets.commit.required.acks=-1 \
          --override offsets.commit.timeout.ms=5000 \
          --override offsets.load.buffer.size=5242880 \
          --override offsets.retention.check.interval.ms=600000 \
          --override offsets.retention.minutes=1440 \
          --override offsets.topic.compression.codec=0 \
          --override offsets.topic.num.partitions=50 \
          --override offsets.topic.replication.factor=3 \
          --override offsets.topic.segment.bytes=104857600 \
          --override queued.max.requests=500 \
          --override quota.consumer.default=9223372036854775807 \
          --override quota.producer.default=9223372036854775807 \
          --override replica.fetch.min.bytes=1 \
          --override replica.fetch.wait.max.ms=500 \
          --override replica.high.watermark.checkpoint.interval.ms=5000 \
          --override replica.lag.time.max.ms=10000 \
          --override replica.socket.receive.buffer.bytes=65536 \
          --override replica.socket.timeout.ms=30000 \
          --override request.timeout.ms=30000 \
          --override socket.receive.buffer.bytes=102400 \
          --override socket.request.max.bytes=104857600 \
          --override socket.send.buffer.bytes=102400 \
          --override unclean.leader.election.enable=true \
          --override zookeeper.session.timeout.ms=6000 \
          --override zookeeper.set.acl=false \
          --override broker.id.generation.enable=true \
          --override connections.max.idle.ms=600000 \
          --override controlled.shutdown.enable=true \
          --override controlled.shutdown.max.retries=3 \
          --override controlled.shutdown.retry.backoff.ms=5000 \
          --override controller.socket.timeout.ms=30000 \
          --override default.replication.factor=1 \
          --override fetch.purgatory.purge.interval.requests=1000 \
          --override group.max.session.timeout.ms=300000 \
          --override group.min.session.timeout.ms=6000 \
          --override inter.broker.protocol.version=2.2.0 \
          --override log.cleaner.backoff.ms=15000 \
          --override log.cleaner.dedupe.buffer.size=134217728 \
          --override log.cleaner.delete.retention.ms=86400000 \
          --override log.cleaner.enable=true \
          --override log.cleaner.io.buffer.load.factor=0.9 \
          --override log.cleaner.io.buffer.size=524288 \
          --override log.cleaner.io.max.bytes.per.second=1.7976931348623157E308 \
          --override log.cleaner.min.cleanable.ratio=0.5 \
          --override log.cleaner.min.compaction.lag.ms=0 \
          --override log.cleaner.threads=1 \
          --override log.cleanup.policy=delete \
          --override log.index.interval.bytes=4096 \
          --override log.index.size.max.bytes=10485760 \
          --override log.message.timestamp.difference.max.ms=9223372036854775807 \
          --override log.message.timestamp.type=CreateTime \
          --override log.preallocate=false \
          --override log.retention.check.interval.ms=300000 \
          --override max.connections.per.ip=2147483647 \
          --override num.partitions=4 \
          --override producer.purgatory.purge.interval.requests=1000 \
          --override replica.fetch.backoff.ms=1000 \
          --override replica.fetch.max.bytes=1048576 \
          --override replica.fetch.response.max.bytes=10485760 \
          --override reserved.broker.max.id=1000 "
        env:
        - name: KAFKA_HEAP_OPTS
          value : "-Xmx512M -Xms512M"
        - name: KAFKA_OPTS
          value: "-Dlogging.level=INFO"
        volumeMounts:
        - name: datadir
          mountPath: /var/lib/kafka
        readinessProbe:
          tcpSocket:
            port: 9092
          timeoutSeconds: 1
          initialDelaySeconds: 5
      securityContext:
        runAsUser: 1000
        fsGroup: 1000
  volumeClaimTemplates:
  - metadata:
      name: datadir
    spec:
      accessModes: [ "ReadWriteMany" ]
      storageClassName: cephfs
      resources:
        requests:
          storage:  20Gi

4.3 部署

這裡zookeeper和kafka都是有狀態的服務，不能使用deployment和rc這種控制器來部署，這裡我們使用statefulset來部署zookeeper和kafka服務。

4.3.1 給節點打標籤

這裡我們想在哪幾臺機器上來執行kafka，需要對節點進行打標籤。

kubectl label node [node-name] travis.io/schedule-only=kafka 

當然，如果我們如果不想在哪些節點執行kafka，可以通過配置汙點來進行。

kubectl taint node [node-name] travis.io/schedule-only=kafka:NoSchedule

4.3.2 建立名稱空間

[root@k8s001 kafka]# kubectl create ns kafka

4.3.3 建立zookeeper服務

# 建立zookeeper服務
[root@k8s001 kafka]# kubectl apply -f zookeeper.yaml
# 檢視zookeeper服務執行狀態
[root@k8s001 kafka]# kubectl get pod -n kafka
NAME      READY   STATUS        RESTARTS   AGE
zk-0      1/1     Running       0          7m8s
zk-1      1/1     Running       0          7m8s
zk-2      1/1     Running       0          7m8s

4.3.4 建立kafka服務

[root@k8s001 kafka]# kubectl apply -f kafka.yaml
[root@k8s001 kafka]# kubectl get pod -n kafka
NAME      READY   STATUS    RESTARTS   AGE
kafka-0   1/1     Running   0          11m
kafka-1   1/1     Running   0          11m
kafka-2   1/1     Running   0          10m
zk-0      1/1     Running   0          6m44s
zk-1      1/1     Running   0          6m44s
zk-2      1/1     Running   0          6m44s

4.3.5 測試

測試zookeeper：
kubectl exec -it zk-0 -n kafka -- zkServer.sh status
kubectl exec -it zk-0 -n kafka -- zkCli.sh create /hello world
kubectl delete -f zookeeper.yaml
kubectl apply -f zookeeper.yaml
kubectl exec -it zk-0 -n kafka -- zkCli.sh get /hello
測試kafka:
kubectl exec -it kafka-0 -n kafka -- bash 
>kafka-topics.sh --create \
--topic test \
--zookeeper zk-0.zk-hs.kafka.svc.cluster.local:2181,zk-1.zk-hs.kafka.svc.cluster.local:2181,zk-2.zk-hs.kafka.svc.cluster.local:2181 \
--partitions 3 \
--replication-factor 2
kafka-topics.sh --list --zookeeper zk-0.zk-hs.kafka.svc.cluster.local:2181,zk-1.zk-hs.kafka.svc.cluster.local:2181,zk-2.zk-hs.kafka.svc.cluster.local:2181
kafka-console-consumer.sh --topic test --bootstrap-server localhost:9092
# 另起一個視窗，進入kafka-1容器
kubectl exec -it kafka-1 -n kafka -- bash
>kafka-console-producer.sh --topic test --broker-list localhost:9092
隨便輸入內容，觀察kafka-0啟動的kafka-console-consumer.sh的輸出。