metrices-server从
api-server
中获取cpu、内存使用率等监控指标
1、收集HPA控制下所有Pod最近的cpu使用情况(CPU utilization)
2、对比在扩容条件里记录的cpu限额(CPUUtilization)
3、调整实例数(必须要满足不超过最大/最小实例数)
4、每隔30s做一次自动扩容的判断
CPU utilization的计算方法是用cpu usage(最近一分钟的平均值,通过metrics可以直接获取到)除以cpu request(这里cpu request就是我们在创建容器时制定的cpu使用核心数)得到一个平均值,这个平均值可以理解为:平均每个Pod CPU核心的使用占比。
k8s中的某个Metrics Server(Heapster或自定义Metrics Server)持续采集所有Pod副本的指标数据。
HPA控制器通过Metrics Server的API(Heapster的API或聚合API)获取这些数据,基于用户定义的扩缩容规则进行计算,得到目标Pod副本数量。
当目标Pod副本数量与当前副本数量不同时,HPA控制器就访问Pod的副本控制器(Deployment 、RC或者ReplicaSet)发起scale操作,调整Pod的副本数量,完成扩缩容操作。
总结:HPA 通过监控分析一些控制器控制的所有 Pod 的负载变化情况来确定是否需要调整 Pod 的副本数量。
Master的kube-controller-manager服务持续监测目标Pod的某种性能指标,以计算是否需要调整副本数量。目前k8s支持的指标类型如下。
◎ Pod资源使用率:Pod级别的性能指标,通常是一个比率值,例如CPU使用率。
◎ Pod自定义指标:Pod级别的性能指标,通常是一个数值,例如接收的请求数量。
◎ Object自定义指标或外部自定义指标:通常是一个数值,需要容器应用以某种方式提供,例如通过HTTP URL“/metrics”提供,或者使用外部服务提供的指标采集URL。
k8s从1.11版本开始,弃用基于Heapster组件完成Pod的CPU使用率采集的机制,全面转向基于Metrics Server完成数据采集。Metrics Server将采集到的Pod性能指标数据通过聚合API(Aggregated API)如metrics.k8s.io、custom.metrics.k8s.io和external.metrics.k8s.io提供给HPA控制器进行查询。关于聚合API和聚合器(API Aggregator)的概念我们后面详细讲解。
通过 伸缩系数 判断是否要进行扩容或缩容。
HPA会根据获得的指标数值,应用相应的算法算出一个伸缩系数,此系数是指标的期望值与目前值的比值,如果大于1表示扩容,小于1表示缩容。
容忍度
--horizontal-pod-autoscaler-tolerance:容忍度
它允许一定范围内的使用量的不稳定,现在默认为0.1,这也是出于维护系统稳定性的考虑。
例如,设定HPA调度策略为cpu使用率高于50%触发扩容,那么只有当使用率大于55%或者小于45%才会触发伸缩活动,HPA会尽力把Pod的使用率控制在这个范围之间。
算法
具体的每次扩容或者缩容的多少Pod的算法为: 期望副本数 = ceil[当前副本数 * ( 当前指标 / 期望指标 )]
举个栗子
当前metric值是200m,期望值是100m,那么pod副本数将会翻一倍,因为 比率为 200.0 / 100.0 = 2.0;
如果当前值是 50m ,我们会将pod副本数减半,因为 50.0 / 100.0 == 0.5
如果比率接近1.0,如0.9或1.1(即容忍度是0.1),将不会进行缩放(取决于内置的全局参数容忍度,–horizontal-pod-autoscaler-tolerance,默认值为0.1)。
此外,存在几种Pod异常的情况,如下所述。
◎ Pod正在被删除(设置了删除时间戳):将不会计入目标Pod副本数量。
◎ Pod的当前指标值无法获得:本次探测不会将这个Pod纳入目标Pod副本数量,后续的探测会被重新纳入计算范围。
◎ 如果指标类型是CPU使用率,则对于正在启动但是还未达到Ready状态的Pod,也暂时不会纳入目标副本数量范围。可以通过kube-controller-manager服务的启动参数--horizontal-pod-autoscaler-initial-readiness-delay设置首次探测Pod是否Ready的延时时间,默认值为30s。另一个启动参数--horizontal-pod-autoscaler-cpuinitialization-period设置首次采集Pod的CPU使用率的延时时间。
注意:
冷却和延迟机制
使用HPA管理一组副本时,有可能因为metrics动态变化而导致副本数频繁波动,这种现象叫做 “颠簸”。
想象一种场景:
当pod所需要的CPU负荷过大,从而在创建一个新pod的过程中,系统的CPU使用量可能会同样在有一个攀升的过程。所以,在每一次作出决策后的一段时间内,将不再进行扩展决策。对于扩容而言,这个时间段为3分钟,缩容为5分钟
Pod延迟探测机制
如果指标类型是CPU使用率,则对于正在启动但是还未达到Ready状态的Pod,也暂时不会纳入目标副本数量范围。
Kubernetes是借助Agrregator APIServer扩展机制来实现Custom Metrics。Custom Metrics APIServer是一个提供查询Metrics指标的API服务(Prometheus的一个适配器),这个服务启动后,kubernetes会暴露一个叫custom.metrics.k8s.io的API,当请求这个URL时,请求通过Custom Metics APIServer去Prometheus里面去查询对应的指标,然后将查询结果按照特定格式返回。
要支持最新的custom(包括external)的metrics,也需要使用新版本的HPA:autoscaling/v2beta1,里面增加四种类型的Metrics:Resource、Pods、Object、External,每种资源对应不同的场景,下面分别说明:
在HPA最新的版本 autoscaling/v2beta2 中又对metrics的配置和HPA扩缩容的策略做了完善,特别是对 metrics 数据的目标指标值的类型定义更通用灵活:包括AverageUtilization、AverageValue和Value,但是不是所有的类型的Metrics都支持三种目标值的,具体对应关系如下表。
HPA里的各种类型的Metrics和Metrics Target Type的对应支持关系表
Metrics Type \ Target Type | AverageUtilization | AverageValue | Value | 备注(query metrics) |
Resource(pod’s cpu/memory etc.) | Yes | Yes | No | pods metrics list |
Pods(pod’s other metrics) | No | Yes | No | pods metrics list |
Object(k8s object) | No | Yes | Yes | object metrics |
External(not k8s object) | No | Yes | Yes | external metrics list |
先看个最简单的HPA的定义的例子
- apiVersion: autoscaling/v2beta2
- kind: HorizontalPodAutoscaler
- metadata:
- name: php-apache
- spec:
- scaleTargetRef:
- apiVersion: apps/v1
- kind: Deployment
- name: php-apache
- minReplicas: 1
- maxReplicas: 10
- metrics:
- - type: Resource
- resource:
- name: cpu
- target:
- type: Utilization
- averageUtilization: 50
从上面的例子可以看出,HPA的spec定义由三个必填部分组成:
完整的HPA的定义可参考k8s的官方API文档。
默认HPA spec里不配置任何metrics的话k8s会默认设置cpu的Resouce,且目标类型是AverageUtilization value为80%。
查看HPA所有版本
- [root@master C]# kubectl api-versions |grep autoscaling
- autoscaling/v1 #只支持通过CPU为参考依据来改变Pod的副本数
- autoscaling/v2beta1 #支持通过CPU、内存、连接数或者自定义规则为参考依据
- autoscaling/v2beta2 #和v2beta1差不多
查看当前版本
- [root@master C]# kubectl explain hpa
- KIND: HorizontalPodAutoscaler
- VERSION: autoscaling/v1 #可以看到使用的默认版本是v1
-
- DESCRIPTION:
- configuration of a horizontal pod autoscaler.
-
- FIELDS:
- apiVersion
- APIVersion defines the versioned schema of this representation of an
- object. Servers should convert recognized schemas to the latest internal
- value, and may reject unrecognized values. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
-
- kind
- Kind is a string value representing the REST resource this object
- represents. Servers may infer this from the endpoint the client submits
- requests to. Cannot be updated. In CamelCase. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
-
- metadata
- Standard object metadata. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#metadata
-
- spec
- behaviour of autoscaler. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#spec-and-status.
-
- status
- current information about the autoscaler.
指定使用版本,这里并不是修改,相当于执行这条命令时,指定了下版本
- [root@master C]# kubectl explain hpa --api-version=autoscaling/v2beta1
- KIND: HorizontalPodAutoscaler
- VERSION: autoscaling/v2beta1
-
- DESCRIPTION:
- HorizontalPodAutoscaler is the configuration for a horizontal pod
- autoscaler, which automatically manages the replica count of any resource
- implementing the scale subresource based on the metrics specified.
-
- FIELDS:
- apiVersion <string>
- APIVersion defines the versioned schema of this representation of an
- object. Servers should convert recognized schemas to the latest internal
- value, and may reject unrecognized values. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
-
- kind <string>
- Kind is a string value representing the REST resource this object
- represents. Servers may infer this from the endpoint the client submits
- requests to. Cannot be updated. In CamelCase. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
-
- metadata
- metadata is the standard object metadata. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#metadata
-
- spec
- spec is the specification for the behaviour of the autoscaler. More info:
- https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#spec-and-status.
-
- status
- status is the current information about the autoscaler.
- [root@master kube-system]# kubectl top nodes #查看节点状态,因为没有安装,所以会报错
- Error from server (NotFound): the server could not find the requested resource (get services http:heapster:)
- [root@master kube-system]# vim components-v0.5.0.yaml
- apiVersion: v1
- kind: ServiceAccount
- metadata:
- labels:
- k8s-app: metrics-server
- name: metrics-server
- namespace: kube-system
- ---
- apiVersion: rbac.authorization.k8s.io/v1
- kind: ClusterRole
- metadata:
- labels:
- k8s-app: metrics-server
- rbac.authorization.k8s.io/aggregate-to-admin: "true"
- rbac.authorization.k8s.io/aggregate-to-edit: "true"
- rbac.authorization.k8s.io/aggregate-to-view: "true"
- name: system:aggregated-metrics-reader
- rules:
- - apiGroups:
- - metrics.k8s.io
- resources:
- - pods
- - nodes
- verbs:
- - get
- - list
- - watch
- ---
- apiVersion: rbac.authorization.k8s.io/v1
- kind: ClusterRole
- metadata:
- labels:
- k8s-app: metrics-server
- name: system:metrics-server
- rules:
- - apiGroups:
- - ""
- resources:
- - pods
- - nodes
- - nodes/stats
- - namespaces
- - configmaps
- verbs:
- - get
- - list
- - watch
- ---
- apiVersion: rbac.authorization.k8s.io/v1
- kind: RoleBinding
- metadata:
- labels:
- k8s-app: metrics-server
- name: metrics-server-auth-reader
- namespace: kube-system
- roleRef:
- apiGroup: rbac.authorization.k8s.io
- kind: Role
- name: extension-apiserver-authentication-reader
- subjects:
- - kind: ServiceAccount
- name: metrics-server
- namespace: kube-system
- ---
- apiVersion: rbac.authorization.k8s.io/v1
- kind: ClusterRoleBinding
- metadata:
- labels:
- k8s-app: metrics-server
- name: metrics-server:system:auth-delegator
- roleRef:
- apiGroup: rbac.authorization.k8s.io
- kind: ClusterRole
- name: system:auth-delegator
- subjects:
- - kind: ServiceAccount
- name: metrics-server
- namespace: kube-system
- ---
- apiVersion: rbac.authorization.k8s.io/v1
- kind: ClusterRoleBinding
- metadata:
- labels:
- k8s-app: metrics-server
- name: system:metrics-server
- roleRef:
- apiGroup: rbac.authorization.k8s.io
- kind: ClusterRole
- name: system:metrics-server
- subjects:
- - kind: ServiceAccount
- name: metrics-server
- namespace: kube-system
- ---
- apiVersion: v1
- kind: Service
- metadata:
- labels:
- k8s-app: metrics-server
- name: metrics-server
- namespace: kube-system
- spec:
- ports:
- - name: https
- port: 443
- protocol: TCP
- targetPort: https
- selector:
- k8s-app: metrics-server
- ---
- apiVersion: apps/v1
- kind: Deployment
- metadata:
- labels:
- k8s-app: metrics-server
- name: metrics-server
- namespace: kube-system
- spec:
- selector:
- matchLabels:
- k8s-app: metrics-server
- strategy:
- rollingUpdate:
- maxUnavailable: 0
- template:
- metadata:
- labels:
- k8s-app: metrics-server
- spec:
- containers:
- - args:
- - --cert-dir=/tmp
- - --secure-port=4443
- - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
- - --kubelet-use-node-status-port
- - --metric-resolution=15s
- - --kubelet-insecure-tls
- image: registry.cn-shenzhen.aliyuncs.com/zengfengjin/metrics-server:v0.5.0
- imagePullPolicy: IfNotPresent
- livenessProbe:
- failureThreshold: 3
- httpGet:
- path: /livez
- port: https
- scheme: HTTPS
- periodSeconds: 10
- name: metrics-server
- ports:
- - containerPort: 4443
- name: https
- protocol: TCP
- readinessProbe:
- failureThreshold: 3
- httpGet:
- path: /readyz
- port: https
- scheme: HTTPS
- initialDelaySeconds: 20
- periodSeconds: 10
- resources:
- requests:
- cpu: 100m
- memory: 200Mi
- securityContext:
- readOnlyRootFilesystem: true
- runAsNonRoot: true
- runAsUser: 1000
- volumeMounts:
- - mountPath: /tmp
- name: tmp-dir
- nodeSelector:
- kubernetes.io/os: linux
- priorityClassName: system-cluster-critical
- serviceAccountName: metrics-server
- volumes:
- - emptyDir: {}
- name: tmp-dir
- ---
- apiVersion: apiregistration.k8s.io/v1
- kind: APIService
- metadata:
- labels:
- k8s-app: metrics-server
- name: v1beta1.metrics.k8s.io
- spec:
- group: metrics.k8s.io
- groupPriorityMinimum: 100
- insecureSkipTLSVerify: true
- service:
- name: metrics-server
- namespace: kube-system
- version: v1beta1
- versionPriority: 100
- [root@master kube-system]# kubectl apply -f components-v0.5.0.yaml
- serviceaccount/metrics-server created
- clusterrole.rbac.authorization.k8s.io/system:aggregated-metrics-reader created
- clusterrole.rbac.authorization.k8s.io/system:metrics-server created
- rolebinding.rbac.authorization.k8s.io/metrics-server-auth-reader created
- clusterrolebinding.rbac.authorization.k8s.io/metrics-server:system:auth-delegator created
- clusterrolebinding.rbac.authorization.k8s.io/system:metrics-server created
- service/metrics-server created
- deployment.apps/metrics-server created
- apiservice.apiregistration.k8s.io/v1beta1.metrics.k8s.io created
-
- #查看创建的pod
- [root@master kube-system]# kubectl get pods -n kube-system| egrep 'NAME|metrics-server'
- NAME READY STATUS RESTARTS AGE
- metrics-server-5944675dfb-q6cdd 0/1 ContainerCreating 0 6s
-
- #查看日志
- [root@master kube-system]# kubectl logs metrics-server-5944675dfb-q6cdd -n kube-system
- I0718 03:06:39.064633 1 serving.go:341] Generated self-signed cert (/tmp/apiserver.crt, /tmp/apiserver.key)
- I0718 03:06:39.870097 1 configmap_cafile_content.go:202] Starting client-ca::kube-system::extension-apiserver-authentication::client-ca-file
- I0718 03:06:39.870122 1 configmap_cafile_content.go:202] Starting client-ca::kube-system::extension-apiserver-authentication::requestheader-client-ca-file
- I0718 03:06:39.870159 1 shared_informer.go:240] Waiting for caches to sync for client-ca::kube-system::extension-apiserver-authentication::client-ca-file
- I0718 03:06:39.870160 1 shared_informer.go:240] Waiting for caches to sync for client-ca::kube-system::extension-apiserver-authentication::requestheader-client-ca-file
- I0718 03:06:39.870105 1 requestheader_controller.go:169] Starting RequestHeaderAuthRequestController
- I0718 03:06:39.871166 1 shared_informer.go:240] Waiting for caches to sync for RequestHeaderAuthRequestController
- I0718 03:06:39.872804 1 dynamic_serving_content.go:130] Starting serving-cert::/tmp/apiserver.crt::/tmp/apiserver.key
- I0718 03:06:39.875741 1 secure_serving.go:197] Serving securely on [::]:4443
- I0718 03:06:39.876050 1 tlsconfig.go:240] Starting DynamicServingCertificateController
- I0718 03:06:39.970469 1 shared_informer.go:247] Caches are synced for client-ca::kube-system::extension-apiserver-authentication::client-ca-file
- I0718 03:06:39.970575 1 shared_informer.go:247] Caches are synced for client-ca::kube-system::extension-apiserver-authentication::requestheader-client-ca-file
- I0718 03:06:39.971610 1 shared_informer.go:247] Caches are synced for RequestHeaderAuthRequestController
-
- #如果报错的化,可以修改apiserver的yaml文件,这是k8s的yaml文件
- [root@master kube-system]# vim /etc/kubernetes/manifests/kube-apiserver.yaml
-
- 40 - --tls-cert-file=/etc/kubernetes/pki/apiserver.crt
- 41 - --tls-private-key-file=/etc/kubernetes/pki/apiserver.key
- 42 - --enable-aggregator-routing=true #添加这行
- 43 image: registry.aliyuncs.com/google_containers/kube-apiserver:v1.18.0
- 44 imagePullPolicy: IfNotPresent
-
- #保存退出
- [root@master kube-system]# systemctl restart kubelet #修改后重启kubelet
-
- #再次查看节点信息
-
- [root@master kube-system]# kubectl top node
- NAME CPU(cores) CPU% MEMORY(bytes) MEMORY%
- master 327m 4% 3909Mi 23%
- node 148m 1% 1327Mi 8%
- [root@master test]# cat nginx.yaml
- apiVersion: apps/v1
- kind: Deployment
- metadata:
- name: nginx
- spec:
- selector:
- matchLabels:
- run: nginx
- replicas: 1
- template:
- metadata:
- labels:
- run: nginx
- spec:
- containers:
- - name: nginx
- image: nginx:1.15.2
- ports:
- - containerPort: 80
- resources:
- limits:
- cpu: 500m
- requests: #想要HPA生效,必须添加requests声明
- cpu: 200m
-
- ---
- apiVersion: v1
- kind: Service
- metadata:
- name: nginx
- labels:
- run: nginx
- spec:
- ports:
- - port: 80
- selector:
- run: nginx
- [root@master test]# kubectl get pods -o wide
- NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
- nginx-9cb8d65b5-tq9v4 1/1 Running 0 14m 10.244.1.22 node <none> <none>
- [root@master test]# kubectl get svc nginx
- NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
- nginx ClusterIP 172.16.169.27 <none> 80/TCP 15m
- [root@master test]# kubectl describe svc nginx
- Name: nginx
- Namespace: default
- Labels: run=nginx
- Annotations: Selector: run=nginx
- Type: ClusterIP
- IP: 172.16.169.27
- Port: <unset> 80/TCP
- TargetPort: 80/TCP
- Endpoints: 10.244.1.22:80
- Session Affinity: None
- Events: <none>
- [root@node test]# curl 172.16.169.27 #访问成功
- html>
- <html>
- <head>
- <title>Welcome to nginx!title>
- <style>
- body {
- width: 35em;
- margin: 0 auto;
- font-family: Tahoma, Verdana, Arial, sans-serif;
- }
- style>
- head>
- <body>
- <h1>Welcome to nginx!h1>
- <p>If you see this page, the nginx web server is successfully installed and
- working. Further configuration is required.p>
-
- <p>For online documentation and support please refer to
- <a href="http://nginx.org/">nginx.orga>.<br/>
- Commercial support is available at
- <a href="http://nginx.com/">nginx.coma>.p>
-
- <p><em>Thank you for using nginx.em>p>
- body>
- html>
- #创建一个cpu利用率达到20,最大10个pod,最小1个,这里没有指定版本所以默认是v1版本,而v1版本只能以CPU为标准
- [root@master test]# kubectl autoscale deployment nginx --cpu-percent=20 --min=1 --max=10
- horizontalpodautoscaler.autoscaling/nginx autoscaled
-
- #TARGETS可以看到使用率
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx Deployment/nginx 0%/20% 1 10 1 86s
-
- #创建一个测试pod增加负载,访问地址要和pod的svc地址相同
- [root@master ~]# kubectl run busybox -it --image=busybox -- /bin/sh -c 'while true; do wget -q -O- http://10.244.1.22; done'
-
-
- #过一分钟后看hap的使用率,REPLICAS是当前pod的数量
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx Deployment/nginx 27%/20% 1 10 5 54m
-
- [root@master test]# kubectl get pods #再看pod数量,发现已经增加到了5个
- NAME READY STATUS RESTARTS AGE
- bustbox 1/1 Running 0 119s
- nginx-9cb8d65b5-24dg2 1/1 Running 0 57s
- nginx-9cb8d65b5-c6n98 1/1 Running 0 87s
- nginx-9cb8d65b5-ksjzv 1/1 Running 0 57s
- nginx-9cb8d65b5-n77fm 1/1 Running 0 87s
- nginx-9cb8d65b5-tq9v4 1/1 Running 0 84m
- [root@master test]# kubectl get deployments.apps
- NAME READY UP-TO-DATE AVAILABLE AGE
- nginx 5/5 5 5 84m
-
-
- #此时,停止压测,过好几分钟后再次查看pod数量和使用率
- [root@master test]# kubectl delete pod busybox #终止后,删除pod
- [root@master test]# kubectl get hpa #虽然使用率已经降到0了,但是可以看到当前REPLICAS的数量还5,这个需要等一会就会缩容
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx Deployment/nginx 0%/20% 1 10 5 58m
-
- #过了几分钟后,可以看到pod数量已经回到了1
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx Deployment/nginx 0%/20% 1 10 1 64m
- [root@master test]# kubectl get pods
- NAME READY STATUS RESTARTS AGE
- nginx-9cb8d65b5-tq9v4 1/1 Running 0 95m
- #先把上面创建的资源删除
- [root@master test]# kubectl delete horizontalpodautoscalers.autoscaling nginx
- horizontalpodautoscaler.autoscaling "nginx" deleted
- [root@master test]# kubectl delete -f nginx.yaml
- deployment.apps "nginx" deleted
- service "nginx" deleted
- [root@master test]# cat nginx.yaml
- apiVersion: apps/v1
- kind: Deployment
- metadata:
- name: nginx
- spec:
- selector:
- matchLabels:
- run: nginx
- replicas: 1
- template:
- metadata:
- labels:
- run: nginx
- spec:
- containers:
- - name: nginx
- image: nginx:1.15.2
- ports:
- - containerPort: 80
- resources:
- limits:
- cpu: 500m
- memory: 60Mi
- requests:
- cpu: 200m
- memory: 25Mi
-
- ---
- apiVersion: v1
- kind: Service
- metadata:
- name: nginx
- labels:
- run: nginx
- spec:
- ports:
- - port: 80
- selector:
- run: nginx
-
- [root@master test]# kubectl apply -f nginx.yaml
- deployment.apps/nginx created
- service/nginx created
- [root@master test]# vim hpa-nginx.yaml
- apiVersion: autoscaling/v2beta1 #上面的hpa版本有提到过,使用基于内存的hpa需要换个版本
- kind: HorizontalPodAutoscaler
- metadata:
- name: nginx-hpa
- spec:
- maxReplicas: 10 #1-10的pod数量限制
- minReplicas: 1
- scaleTargetRef: #指定使用hpa的资源对象,版本、类型、名称要和上面创建的相同
- apiVersion: apps/v1
- kind: Deployment
- name: nginx
- metrics:
- - type: Resource
- resource:
- name: memory
- targetAverageUtilization: 50 #限制%50的内存
- [root@master test]# kubectl apply -f hpa-nginx.yaml
- horizontalpodautoscaler.autoscaling/nginx-hpa created
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx-hpa Deployment/nginx 7%/50% 1 10 1 59s
- #在pod中执行命令,增加内存负载
- [root@master ~]# kubectl exec -it nginx-78f4944bb8-2rz7j -- /bin/sh -c 'dd if=/dev/zero of=/tmp/file1'
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx-hpa Deployment/nginx 137%/50% 1 10 1 12m
- [root@master test]# kubectl get hpa
- NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS AGE
- nginx-hpa Deployment/nginx 14%/50% 1 10 3 12m
- [root@master test]# kubectl get pods
- NAME READY STATUS RESTARTS AGE
- nginx-78f4944bb8-2rz7j 1/1 Running 0 21m
- nginx-78f4944bb8-bxh78 1/1 Running 0 34s
- nginx-78f4944bb8-g8w2h 1/1 Running 0 34s
- #与CPU相同,内存上去了也会自动创建pod