玩k8s? Pod的资源限制&重启策略

最新推荐文章于 2025-01-19 22:57:29 发布

原创

最新推荐文章于 2025-01-19 22:57:29 发布 · 1.9k 阅读

2 ·

CC 4.0 BY-SA版权

文章标签：

#kubernetes #docker #pod资源

本文介绍了Kubernetes Pod的资源限制，包括内存和CPU的申请配额与限制配额，以及资源限制的重要性。同时，讲解了Pod的重启策略，如Always、OnFailure和Never，并给出了不同场景下的适用情况。最后，详细阐述了健康检查机制，包括livenessProbe和readinessProbe，以及httpGet、exec和tcpSocket三种检查方法的应用示例。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

Pod 资源限制&重启策略

文章目录

Pod 资源限制&重启策略

1.1: 资源限制

官方文档

https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/

Pod资源配额有两种：

申请配额：调度时使用，参考是否有节点满足该配置

spec.containers[].resources.limits.cpu

spec.containers[].resources.limits.memory
限制配额：容器能使用的最大配置

spec.containers[].resources.requests.cpu

spec.containers[].resources.requests.memory
容器资源限制：
1、内存
2、CPU

limits：实际最大使用的配额

requests：申请的配额，主要用于k8s做资源调度分配时参考值

一般limits比requests大一些
资源限制好处：避免某容器资源利用率异常突发影响其他容器，可能会产生雪崩效应！

  - name: wp
    image: wordpress
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"

示例：

[root@k8s-master1 demo]# vim resources.yaml
apiVersion: v1
kind: Pod
metadata:
  name: frontend
spec:
  containers:
  - name: db
    image: mysql
    env: "设定变量"
    - name: MYSQL_ROOT_PASSWORD
      value: "password"
    resources: "资源限额"
      requests: "请求资源"
        memory: "64Mi" "请求内存64M"
        cpu: "250m" "一秒钟能够使用一个cpu核心资源的25%的资源"
      limits: "最大限制资源"
        memory: "128Mi" "最大限制内存128M"
        cpu: "500m" "一秒钟能够使用一个cpu核心资源的50%资源"
  - name: wp
    image: wordpress
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"
[root@k8s-master1 demo]# kubectl apply -f resources.yaml 
pod/frontend created
[root@k8s-master1 demo]# kubectl describe pod  frontend "查看创建过程"
[root@k8s-master1 demo]# kubectl get pods
NAME               READY   STATUS    RESTARTS   AGE
frontend           2/2     Running   0          2m39s
[root@k8s-master1 demo]# kubectl get pods -o wide
NAME       READY   STATUS    RESTARTS   AGE     IP            NODE        NOMINATED NODE   READINESS GATES
frontend   2/2     Running   0          3m55s   10.244.1.33   k8s-node1   <none>           <none>
[root@k8s-master1 demo]#  kubectl describe nodes k8s-node1
...
  Namespace                   Name                                          CPU Requests  CPU Limits  Memory Requests  Memory Limits  AGE
  ---------                   ----                                          ------------  ----------  ---------------  -------------  ---
  default                     frontend                                      500m (12%)    1 (25%)     128Mi (7%)       256Mi (14%)    4m42s
...
Allocated resources:
  (Total limits may be over 100 percent, i.e., overcommitted.)
  Resource           Requests      Limits
  --------           --------      ------
  cpu                920m (23%)    1320m (33%)
  memory             1308Mi (76%)  1878Mi (109%)
  ephemeral-storage  0 (0%)        0 (0%)
  hugepages-1Gi      0 (0%)        0 (0%)
  hugepages-2Mi      0 (0%)        0 (0%)
Events:              <none>
[root@k8s-master1 demo]# kubectl get pods
NAME       READY   STATUS    RESTARTS   AGE
frontend   2/2     Running   0          6m9s
[root@k8s-master1 demo]# kubectl describe nodes 192.168.100.202 "查看节点信息"

[root@k8s-master1 demo]# kubectl get ns "查看命名空间"
NAME                   STATUS   AGE
default                Active   3d1h
ingress-nginx          Active   3d1h
kube-node-lease        Active   3d1h
kube-public            Active

最低0.47元/天解锁文章