22,Horizontal Pod Autoscaler(HPA)，自动扩缩容

最新推荐文章于 2022-07-04 17:48:15 发布

安達と島村

最新推荐文章于 2022-07-04 17:48:15 发布

阅读量245

点赞数

分类专栏： k8s 文章标签： python docker

本文链接：https://blog.csdn.net/weixin_43292547/article/details/120771826

版权

k8s 专栏收录该内容

38 篇文章 9 订阅

订阅专栏

在前面的课程中，我们已经可以实现通过手工执行kubectl scale命令实现Pod扩容或缩容，但是这显然不符合Kubernetes的定位目标–自动化、智能化。 Kubernetes期望可以实现通过监测Pod的使用情况，实现pod数量的自动调整，于是就产生了Horizontal Pod Autoscaler（HPA）这种控制器。

HPA可以获取每个Pod利用率，然后和HPA中定义的指标进行对比，同时计算出需要伸缩的具体值，最后实现Pod的数量的调整。其实HPA与之前的Deployment一样，也属于一种Kubernetes资源对象，它通过追踪分析RC控制的所有目标Pod的负载变化情况，来确定是否需要针对性地调整目标Pod的副本数，这是HPA的实现原理。
接下来，我们来做一个实验

1,安装metrics-server

metrics-server可以用来收集集群中的资源使用情况

安装git
yum install git -y

获取metrics-server, 注意使用的版本
git clone -b v0.3.6 https://github.com/kubernetes-incubator/metrics-server

修改deployment, 注意修改的是镜像和初始化参数
cd /root/metrics-server/deploy/1.8+/
vim metrics-server-deployment.yaml

添加/修改下面选项
hostNetwork: true
image: registry.cn-hangzhou.aliyuncs.com/google_containers/metrics-server-amd64:v0.3.6
args:
- --kubelet-insecure-tls
- --kubelet-preferred-address-types=InternalIP,Hostname,InternalDNS,ExternalDNS,ExternalIP
'''添加后文件部分内容如下
    spec:
      hostNetwork: true
      serviceAccountName: metrics-server
      volumes:
      # mount in tmp so we can safely use from-scratch images and/or read-only containers
      - name: tmp-dir
        emptyDir: {}
      containers:
      - name: metrics-server
        image: registry.cn-hangzhou.aliyuncs.com/google_containers/metrics-server-amd64:v0.3.6
        imagePullPolicy: Always
        args:
        - --kubelet-insecure-tls
        - --kubelet-preferred-address-types=InternalIP,Hostname,InternalDNS,ExternalDNS,ExternalIP
        volumeMounts:
        - name: tmp-dir
          mountPath: /tmp
'''

安装metrics-server，这里apply /root/metrics-server/deploy/1.8+/ 里所有的配置文件
kubectl apply -f ./

# 查看metrics-server 的pod的运行情况
kubectl get pod -n kube-system
metrics-server-6b976979db-2xwbj   1/1     Running   0          90s

# 使用kubectl top 资源名   查看资源使用情况，采集信息需要一定时间
kubectl top node
NAME     CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%   
master   83m          4%     1027Mi          59%       
node1    25m          1%     671Mi           38%       
node2    28m          1%     648Mi           37%   

kubectl top pod -n kube-system
NAME                              CPU(cores)   MEMORY(bytes)
coredns-6955765f44-7ptsb          3m           9Mi
coredns-6955765f44-vcwr5          3m           8Mi
etcd-master                       14m          145Mi
metrics-server-6b976979db-dpwrm   1m           11Mi  

至此,metrics-server安装完成

2，部署deployment，servie，HPA

deploy的template的spec里，必须设置resources的limits和requests(貌似设置一个就行)，
否则 kubectl get hpa -n dev查看当前cpu使用状况时，TARGETS显示unknown/3%

cat >pod-test.yaml <<EOF
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deploy
  namespace: dev
spec:
  strategy: # 策略
    type: RollingUpdate  # 滚动更新策略
  replicas: 1
  selector:
    matchLabels:
      app: nginx-pod
  template:
    metadata:
      labels:
        app: nginx-pod
    spec:
      containers:
      - name: nginx
        image: nginx:1.17.1
        resources: # 资源配额
          limits:  # 限制资源（上限）
            cpu: "1" # CPU限制，单位是core数
          requests: # 请求资源（下限）
            cpu: "100m"  # CPU限制，单位是core数
			#如果不设置， kubectl get hpa -n dev的TARGETS显示unknown/3%
			
---

apiVersion: v1
kind: Service
metadata:
    name: nginx-svc
    namespace: dev
spec:
    type: NodePort
    ports:
    - port: 80				
      targetPort: 80
      nodePort: 30000
    selector:
        app: nginx-pod    #根据此标签选择器,选择pod进行expose

---

apiVersion: autoscaling/v1
kind: HorizontalPodAutoscaler
metadata:
  name: pc-hpa
  namespace: dev
spec:
  minReplicas: 1  #最小pod数量
  maxReplicas: 10 #最大pod数量
  targetCPUUtilizationPercentage: 3 # CPU使用率指标
  scaleTargetRef:   # 指定要控制的nginx信息
    apiVersion:  apps/v1
    kind: Deployment
    name: nginx-deploy		
EOF

写入配置文件后apply即可

3，测试

'''
用3个shell分别监控hpa,deploy,pod
'''
kubectl get hpa -n dev -w
kubectl get deploy -n dev -w
kubectl get pod -n dev -w

'''开始测试，用python运行，url里ip和port都是自己设置的'''
import requests
for i in range(10000):
    requests.get('http://192.168.10.100:30000')

kubectl get hpa -n dev -w
'''
NAME     REFERENCE                 TARGETS   MINPODS   MAXPODS   REPLICAS   AGE
pc-hpa   Deployment/nginx-deploy   0%/3%     1         10        1          7m32s
pc-hpa   Deployment/nginx-deploy   1%/3%     1         10        1          8m20s
pc-hpa   Deployment/nginx-deploy   54%/3%    1         10        1          9m21s
pc-hpa   Deployment/nginx-deploy   54%/3%    1         10        4          9m36s
pc-hpa   Deployment/nginx-deploy   54%/3%    1         10        8          9m51s
pc-hpa   Deployment/nginx-deploy   54%/3%    1         10        10         10m
pc-hpa   Deployment/nginx-deploy   0%/3%     1         10        10         10m
pc-hpa   Deployment/nginx-deploy   0%/3%     1         10        10         15m
pc-hpa   Deployment/nginx-deploy   0%/3%     1         10        1          15m
'''

kubectl get deploy -n dev -w
'''
NAME           READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deploy   1/1     1            1           7m23s
nginx-deploy   1/4     1            1           9m21s
nginx-deploy   1/4     4            1           9m21s
nginx-deploy   2/4     4            2           9m22s
nginx-deploy   3/4     4            3           9m23s
nginx-deploy   4/4     4            4           9m23s
nginx-deploy   4/8     4            4           9m36s
nginx-deploy   4/8     8            4           9m36s
nginx-deploy   5/8     8            5           9m37s
nginx-deploy   6/8     8            6           9m38s
nginx-deploy   7/8     8            7           9m38s
nginx-deploy   8/8     8            8           9m38s
nginx-deploy   8/10    8            8           9m51s
nginx-deploy   8/10    10           8           9m51s
nginx-deploy   9/10    10           9           9m52s
nginx-deploy   10/10   10           10          9m53s
nginx-deploy   10/1    10           10          15m
nginx-deploy   1/1     1            1           15m
'''

kubectl get pod -n dev -w
'''
NAME                            READY   STATUS    RESTARTS   AGE
nginx-deploy-547dd76d58-j8lm5   1/1     Running   0          7m26s
nginx-deploy-547dd76d58-wf5gj   0/1     Pending   0          0s
nginx-deploy-547dd76d58-wf5gj   0/1     Pending   0          0s
nginx-deploy-547dd76d58-zd5xr   0/1     Pending   0          0s
nginx-deploy-547dd76d58-cdmqq   0/1     Pending   0          0s
nginx-deploy-547dd76d58-cdmqq   0/1     Pending   0          0s
nginx-deploy-547dd76d58-zd5xr   0/1     Pending   0          0s
nginx-deploy-547dd76d58-wf5gj   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-zd5xr   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-cdmqq   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-zd5xr   1/1     Running             0          1s
nginx-deploy-547dd76d58-wf5gj   1/1     Running             0          2s
nginx-deploy-547dd76d58-cdmqq   1/1     Running             0          2s
nginx-deploy-547dd76d58-mk5rx   0/1     Pending             0          0s
nginx-deploy-547dd76d58-mk5rx   0/1     Pending             0          0s
nginx-deploy-547dd76d58-99kx6   0/1     Pending             0          0s
nginx-deploy-547dd76d58-fl99x   0/1     Pending             0          0s
nginx-deploy-547dd76d58-99kx6   0/1     Pending             0          0s
nginx-deploy-547dd76d58-mk5rx   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-v5kvk   0/1     Pending             0          0s
nginx-deploy-547dd76d58-fl99x   0/1     Pending             0          0s
nginx-deploy-547dd76d58-v5kvk   0/1     Pending             0          0s
nginx-deploy-547dd76d58-fl99x   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-99kx6   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-v5kvk   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-99kx6   1/1     Running             0          1s
nginx-deploy-547dd76d58-fl99x   1/1     Running             0          2s
nginx-deploy-547dd76d58-mk5rx   1/1     Running             0          2s
nginx-deploy-547dd76d58-v5kvk   1/1     Running             0          2s
nginx-deploy-547dd76d58-zw2l2   0/1     Pending             0          0s
nginx-deploy-547dd76d58-zw2l2   0/1     Pending             0          0s
nginx-deploy-547dd76d58-qjhqg   0/1     Pending             0          0s
nginx-deploy-547dd76d58-qjhqg   0/1     Pending             0          0s
nginx-deploy-547dd76d58-zw2l2   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-qjhqg   0/1     ContainerCreating   0          0s
nginx-deploy-547dd76d58-zw2l2   1/1     Running             0          1s
nginx-deploy-547dd76d58-qjhqg   1/1     Running             0          2s
nginx-deploy-547dd76d58-zd5xr   1/1     Terminating         0          5m49s
******
nginx-deploy-547dd76d58-v5kvk   0/1     Terminating         0          5m45s
nginx-deploy-547dd76d58-cdmqq   0/1     Terminating         0          6m
nginx-deploy-547dd76d58-cdmqq   0/1     Terminating         0          6m
'''

可以发现发送大量请求后，pod数自动增加至最大(10)，停止发送后，一定时间后减少到最小(1)