标签：

写在前面

上一篇文章中kubernetes系列教程（五）深入掌握核心观点pod初阶介绍了yaml学习kubernetes中重要的一个观点pod，接下来介绍kubernetes系列教程pod的resource资源打点和pod的Quality of service处事质量。

1. Pod资源打点 1.1 resource界说

容器运行过程中需要分配所需的资源，如何与cggroup联动共同呢？答案是通过界说resource来实现资源的分配，资源的分配单位主要是cpu和memory，资源的界说分两种：requests和limits，requests暗示请求资源，主要用于初始kubernetes调理pod时的依据，暗示必需满足的分配资源;limits暗示资源的限制，即pod不能赶过limits界说的限制巨细，赶过则通过cggroup限制，pod中界说资源可以通过下面四个字段界说：

spec.container[].resources.requests.cpu 请求cpu资源的巨细，如0.1个cpu和100m暗示分配1/10个cpu；

spec.container[].resources.requests.memory 请求内存巨细，单位可用M，Mi，G，，Gi暗示；

spec.container[].resources.limits.cpu 限制cpu的巨细，不能赶过阀值，cggroup中限制的值；

spec.container[].resources.limits.memory 限制内存的巨细，不能赶过阀值，赶过会产生OOM；

1、开始学习如何界说pod的resource资源,如下以界说nginx-demo为例，容器请求cpu资源为250m，限制为500m，请求内存资源为128Mi，限制内存资源为256Mi，固然也可以界说多个容器的资源，多个容器相加就是pod的资源总资源，如下：

[[email protected] demo]#cat nginx-resource.yaml apiVersion: v1 kind: Pod metadata: name: nginx-demo labels: name: nginx-demo spec: containers: - name: nginx-demo image: nginx:1.7.9 imagePullPolicy: IfNotPresent ports: - name: nginx-port-80 protocol: TCP containerPort: 80 resources: requests: cpu: 0.25 memory: 128Mi limits: cpu: 500m memory: 256Mi

2、应用pod的配置界说(如之前的pod还存在，先将其删除kubectl delete pod <pod-name>),或pod定名为此外一个名

[[email protected] demo]# kubectl apply -f nginx-resource.yaml pod/nginx-demo created

3、检察pod资源的分配详情

[[email protected] demo]# kubectl get pods NAME READY STATUS RESTARTS AGE demo-7b86696648-8bq7h 1/1 Running 0 12d demo-7b86696648-8qp46 1/1 Running 0 12d demo-7b86696648-d6hfw 1/1 Running 0 12d nginx-demo 1/1 Running 0 94s [[email protected] demo]# kubectl describe pods nginx-demo Name: nginx-demo Namespace: default Priority: 0 Node: node-3/10.254.100.103 Start Time: Sat, 28 Sep 2019 12:10:49 +0800 Labels: name=nginx-demo Annotations: kubectl.kubernetes.io/last-applied-configuration: {"apiVersion":"v1","kind":"Pod","metadata":{"annotations":{},"labels":{"name":"nginx-demo"},"name":"nginx-demo","namespace":"default"},"sp... Status: Running IP: 10.244.2.13 Containers: nginx-demo: Container ID: docker://55d28fdc992331c5c58a51154cd072cd6ae37e03e05ae829a97129f85eb5ed79 Image: nginx:1.7.9 Image ID: docker-pullable://[email protected]:e3456c851a152494c3e4ff5fcc26f240206abac0c9d794affb40e0714846c451 Port: 80/TCP Host Port: 0/TCP State: Running Started: Sat, 28 Sep 2019 12:10:51 +0800 Ready: True Restart Count: 0 Limits: #限制资源 cpu: 500m memory: 256Mi Requests: #请求资源 cpu: 250m memory: 128Mi Environment: <none> ...省略...

4、Pod的资源如何分配呢？毫无疑问是从node上分配的，当我们创建一个pod的时候如果设置了requests，kubernetes的调理器kube-scheduler会执行两个调理过程：filter过滤和weight称重，kube-scheduler会按照请求的资源过滤，把切合条件的node筛选出来，然后再进行排序，把最满足运行pod的node筛选出来，然后再特定的node上运行pod。调理算法和细节可以参考下kubernetes调理算法介绍。如下是node-3节点资源的分配详情：

[[email protected] ~]# kubectl describe node node-3 ...省略... Capacity: #节点上资源的总资源情况，1个cpu，2g内存，110个pod cpu: 1 ephemeral-storage: 51473888Ki hugepages-2Mi: 0 memory: 1882352Ki pods: 110 Allocatable: #节点容许分配的资源情况，部分预留的资源会排出在Allocatable领域 cpu: 1 ephemeral-storage: 47438335103 hugepages-2Mi: 0 memory: 1779952Ki pods: 110 System Info: Machine ID: 0ea734564f9a4e2881b866b82d679dfc System UUID: FFCD2939-1BF2-4200-B4FD-8822EBFFF904 Boot ID: 293f49fd-8a7c-49e2-8945-7a4addbd88ca Kernel Version: 3.10.0-957.21.3.el7.x86_64 OS Image: CentOS Linux 7 (Core) Operating System: linux Architecture: amd64 Container Runtime Version: docker://18.6.3 Kubelet Version: v1.15.3 Kube-Proxy Version: v1.15.3 PodCIDR: 10.244.2.0/24 Non-terminated Pods: (3 in total) #节点上运行pod的资源的情况，除了nginx-demo之外还有多个pod Namespace Name CPU Requests CPU Limits Memory Requests Memory Limits AGE --------- ---- ------------ ---------- --------------- ------------- --- default nginx-demo 250m (25%) 500m (50%) 128Mi (7%) 256Mi (14%) 63m kube-system kube-flannel-ds-amd64-jp594 100m (10%) 100m (10%) 50Mi (2%) 50Mi (2%) 14d kube-system kube-proxy-mh2gq 0 (0%) 0 (0%) 0 (0%) 0 (0%) 12d Allocated resources: #已经分配的cpu和memory的资源情况 (Total limits may be over 100 percent, i.e., overcommitted.) Resource Requests Limits -------- -------- ------ cpu 350m (35%) 600m (60%) memory 178Mi (10%) 306Mi (17%) ephemeral-storage 0 (0%) 0 (0%) Events: <none> 1.2 资源分配道理