Skywalking8.0.1在K8s中的使用-蒲公英云

Skywalking8.0.1在K8s中的使用

港控/mmm° 2022-11-26 07:47 429阅读 0赞

Skywalking8.0.1在K8s中的使用

1. 准备

ElasticSearch 7.5.1
SkyWalking APM 8.0.1

2. ElasticSearch准备

此处es使用helm3部署，地址 https://github.com/elastic/helm-charts/tree/7.8.1/elasticsearch

clone

将 helm-charts clone 到本地

git clone https://github.com/elastic/helm-charts.git
git checkout 7.8.1

edit

按需修改 helm-charts/elasticsearch/values.yaml 中 imageTag，我需要7.5.1

imageTag: "7.5.1"

install

install:

cd helm-charts
helm install es4skywalking .\es4skywalking\ -n fn-rec

check

查看是否正常启动

kubectl get pods -n <your-namespace> -o wide
kubectl describe pod <pod-name>  -n <your-namespace>

3. skywalking部署

同样通过helm3来安装。

clone

git clone https://github.com/apache/skywalking-kubernetes.git

使用已有的es部署

$ cd chart
$ helm repo add elastic https://helm.elastic.co
$ helm dep up skywalking
$ helm install <release_name> skywalking -n <namespace> \
        --set elasticsearch.enabled=false \
        --set elasticsearch.config.host=<es_host> \
        --set elasticsearch.config.port.http=<es_port> \
        --set elasticsearch.config.user=<es_user> \
        --set elasticsearch.config.password=<es_password>

如：

helm install my-skywalking .\skywalking\ -n fn-rec --set elasticsearch.enabled=false --set elasticsearch.config.host=elasticsearch-skywalking-master-headless --set elasticsearch.config.port.http=9200

check

同样检查是否正确

kubectl get pods -n <your-namespace> -o wide
kubectl describe pod <pod-name>  -n <your-namespace>

4. 项目接入skywalking agent

修改目标项目的dockerFile

FROM maven:3.6.3-openjdk-14 AS MAVEN_BUILD
ARG PROFILE
COPY pom.xml /build/
COPY src /build/src/
WORKDIR /build/
RUN mvn clean package -P${PROFILE} -Dmaven.test.skip=true
FROM adoptopenjdk/openjdk14-openj9
WORKDIR /app
COPY --from=MAVEN_BUILD /build/target/demo.jar /app
# 加javaagent参数很重要
ENTRYPOINT  ["java ", "-javaagent:/sidecar/agent/skywalking-agent.jar", "--enable-preview","-jar", "demo.jar"]

k8s.yaml

注意：镜像中没有JAVA_OPTS这个ENV实际上这个地方不会生效，所以在docker中 -javaagent:/sidecar/agent/skywalking-agent.jar 是需要的。

示例

kind: Deployment
apiVersion: apps/v1
metadata:
  name: search-zd
  labels:
    app: search-zd
spec:
  replicas: 1
  selector:
    matchLabels:
      app: search-zd
  template:
    metadata:
      labels:
        app: search-zd
      annotations:
        prometheus.io/scrape: 'true'
        prometheus.io/path: '/actuator/prometheus'
        prometheus.io/port: '8088'
    spec:
      initContainers:
        - name: skywalking-agent
          #使用本地打包测试镜像，正式环境可以使用dockerhub或者私服的镜像容器中心的容器
          image: 'search-zd/skywalking-agent:v1.8.0.1'
          command: ['cp','-rf','/skywalking/agent/','/sidecar']
          imagePullPolicy: IfNotPresent
          volumeMounts:
            - name: sidecar
              mountPath: /sidecar
      containers:
        - name: search-zd
          image: 'search-zd/fn-es-search:1.0.0.5'
          env:
            # 镜像中没有JAVA_OPTS这个ENV实际上这个地方不会生效，所以在docker中 -javaagent:/sidecar/agent/skywalking-agent.jar 是需要的。
            - name: JAVA_OPTS
              value: -javaagent:/sidecar/agent/skywalking-agent.jar
            - name: SW_AGENT_NAME
              value: search-zd
            - name: SW_AGENT_COLLECTOR_BACKEND_SERVICES
              value: fn-skywalking-oap:11800
          ports:
            - name: server
              containerPort: 8080
              protocol: TCP
            - name: management
              containerPort: 9080
              protocol: TCP
          livenessProbe:
            tcpSocket:
              port: 8080
            timeoutSeconds: 5
          readinessProbe:
            tcpSocket:
              port: 8080
            timeoutSeconds: 5
          resources:
            limits:
              cpu: 2000m
              memory: 4Gi
            requests:
              cpu: 500m
              memory: 2Gi
          imagePullPolicy: IfNotPresent
          volumeMounts:
            - name: sidecar
              mountPath: /sidecar
      restartPolicy: Always
      volumes:
        - name: sidecar
          emptyDir: {}
---
apiVersion: v1
kind: Service
metadata:
  name: search-zd
spec:
  type: ClusterIP
  ports:
    - name: server
      port: 80
      targetPort: 8080
    - name: management
      port: 9080
      targetPort: 9080
  selector:
    app: search-zd

部署

apply:

kubectl apply -f k8s.yaml -n <namespace>

checkout:

kubectl get pods -n <your-namespace> -o wide
kubectl describe pod <pod-name>  -n <your-namespace>

简单查看ui：使用 port-forward 转发一下

kubectl port-forward <target-pod> 8080:8080 -n <namespace>

5. 文档

中文文档：https://github.com/SkyAPM/document-cn-translation-of-skywalking/blob/master/docs/zh/8.0.0/README.md
官网：https://skywalking.apache.org/zh/downloads/