k8s的pod管理

lxw20230271162026-08-26 14:31

一.命令式对象管理

1.命名空间管理

bash 复制代码 
[root@k8s-master ~]# kubectl get namespaces		#查看命名空间
NAME              STATUS   AGE
default           Active   4h20m
kube-flannel      Active   3h31m
kube-node-lease   Active   4h20m
kube-public       Active   4h20m
kube-system       Active   4h20m

#创建命名空间
[root@k8s-master ~]# kubectl get namespaces
NAME              STATUS   AGE
default           Active   4h22m
kube-flannel      Active   3h32m
kube-node-lease   Active   4h22m
kube-public       Active   4h22m
kube-system       Active   4h22m
timinglee         Active   5s

#删除命名空间
[root@k8s-master ~]# kubectl delete namespaces  timinglee
namespace "timinglee" deleted

2.pod管理

bash 复制代码 
#查看pod的运行情况和在哪里运行
[root@k8s-master ~]# kubectl get pods  -o wide
No resources found in default namespace.

#创建pod
[root@k8s-master ~]# kubectl run  lee --image nginx:latest
pod/lee created


[root@k8s-master ~]# kubectl get pods  -o wide
NAME   READY   STATUS    RESTARTS   AGE   IP            NODE        NOMINATED NODE   READINESS GATES
lee    1/1     Running   0          25s   10.244.1.10   k8s-node1   <none>           <none>


#当pod创建出现问题
[root@k8s-master ~]# kubectl run  error --image lee:v1
[root@k8s-master ~]# kubectl get pods -o wide
NAME    READY   STATUS             RESTARTS   AGE   IP            NODE        NOMINATED NODE   READINESS GATES
error   0/1     ImagePullBackOff   0          38s   10.244.2.3    k8s-node2   <none>           <none>
lee     1/1     Running            0          91s   10.244.1.10   k8s-node1   <none>           <none>

#查看pod运行的详细信息
[root@k8s-master ~]# kubectl describe  pods  error
Name:             error
Namespace:        default
Priority:         0
Service Account:  default
Node:             k8s-node2/172.25.254.20
Start Time:       Thu, 20 Aug 2026 15:08:30 +0800
Labels:           run=error
Annotations:      <none>
Status:           Pending
IP:               10.244.2.3
IPs:
  IP:  10.244.2.3
Containers:
  error:
    Container ID:
    Image:          lee:v1
    Image ID:
    Port:           <none>
    Host Port:      <none>
    State:          Waiting
      Reason:       ImagePullBackOff
    Ready:          False
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-kfcjl (ro)
Conditions:
  Type                        Status
  PodReadyToStartContainers   True
  Initialized                 True
  Ready                       False
  ContainersReady             False
  PodScheduled                True
Volumes:
  kube-api-access-kfcjl:
    Type:                    Projected (a volume that contains injected data from multiple sources)
    TokenExpirationSeconds:  3607
    ConfigMapName:           kube-root-ca.crt
    Optional:                false
    DownwardAPI:             true
QoS Class:                   BestEffort
Node-Selectors:              <none>
Tolerations:                 node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                             node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:
  Type     Reason     Age                From               Message
  ----     ------     ----               ----               -------
  Normal   Scheduled  87s                default-scheduler  Successfully assigned default/error to k8s-node2
  Warning  Failed     28s (x2 over 65s)  kubelet            spec.containers{error}: Failed to pull image "lee:v1": Error response from daemon: failed to resolve reference "docker.io/library/lee:v1": docker.io/library/lee:v1: not found
  Warning  Failed     28s (x2 over 65s)  kubelet            spec.containers{error}: Error: ErrImagePull
  Normal   BackOff    16s (x2 over 64s)  kubelet            spec.containers{error}: Back-off pulling image "lee:v1"
  Warning  Failed     16s (x2 over 64s)  kubelet            spec.containers{error}: Error: ImagePullBackOff
  Normal   Pulling    2s (x3 over 86s)   kubelet            spec.containers{error}: Pulling image "lee:v1"
  
#删除  
[root@k8s-master ~]# kubectl delete  pods error
[root@k8s-master ~]# kubectl delete pods --all
pod "error" deleted from default namespace
pod "lee" deleted from default namespace

二.kubectl 命令实操

1.上传实验镜像到仓库library中

bash 复制代码 
[root@k8s-master ~]# docker load  -i myapp.tar.gz
[root@k8s-master ~]# docker tag timinglee/myapp:v1 reg.timinglee.org/library/myapp:v1
[root@k8s-master ~]# docker push  reg.timinglee.org/library/myapp:v1
[root@k8s-master ~]# docker tag timinglee/myapp:v2 reg.timinglee.org/library/myapp:v2
[root@k8s-master ~]# docker push  reg.timinglee.org/library/myapp:v2

2.生成实验所需yml文件

bash 复制代码 
[root@k8s-master ~]# vim replica.yml
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  labels:
    app: replica	#设定控制器标签
  name: replica
spec:
  replicas: 2		#启动pod数量
  selector:
    matchLabels:
      app: replica	#控制器标签选择器
  template:
    metadata:
      labels:
        app: replica	#开启pod的属性模板
    spec:
      containers:
      - image: myapp:v1
        name: myapp

3.kubectl命令使用方法

命令 作用 示例
create 创建资源（通过 yaml 文件或者命令直接创建） kubectl create -f pod.yaml
edit 直接编辑在线运行的资源，修改后即时生效 kubectl edit deployment nginx
patch 打补丁，局部修改资源，不用完整 yaml kubectl patch pod pod-name -p '{"spec":{"...":...}}'
expose 将 Pod/Deployment 暴露为 Service，创建访问入口 kubectl expose deployment nginx --port=80 --type=NodePort
logs 查看 Pod 容器日志，排错最常用 kubectl logs -f pod-name -f实时跟踪日志
attach 附着到 Pod 里面的进程，连接容器（不常用） kubectl attach pod-name
exec 在 Pod 容器内部执行命令，进入容器终端 kubectl exec -it pod-name -- /bin/bash
cp 宿主机和 Pod 容器之间拷贝文件 kubectl cp 本地文件 pod名称:/容器路径
rollout 管理版本滚动更新：查看历史、回滚、重启 kubectl rollout history deployment nginx``kubectl rollout undo deployment nginx回滚上一版本
scale 扩缩副本数，调整 deployment/pod 副本数量 kubectl scale deployment nginx --replicas=3
label 给资源打标签 / 删除标签，用于筛选资源 kubectl label pod pod-name env=test删除标签：kubectl label pod pod-name env-

三.利用控制器实现版本更替

1.建立控制器（在做以下实验时harbor仓库中的library项目中必须有myapp:v1和myapp:v2两个镜像 ）

bash 复制代码 
[root@k8s-master ~]# kubectl create deployment webcluster --image myapp:v1 --replicas 2 --dry-run=client -o yaml > webcluster.yml

root@k8s-master ~]# vim webcluster.yml
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: webcluster
  name: webcluster
spec:
  replicas: 2
  selector:
    matchLabels:
      app: webcluster
  template:
    metadata:
      labels:
        app: webcluster
    spec:
      containers:
      - image: myapp:v1
        name: myapp

[root@k8s-master ~]# kubectl apply -f webcluster.yml
deployment.apps/webcluster created
[root@k8s-master ~]# kubectl get pods
NAME                          READY   STATUS    RESTARTS   AGE
webcluster-77c87d9946-2xl44   1/1     Running   0          3s
webcluster-77c87d9946-cd9zv   1/1     Running   0          3s
[root@k8s-master ~]# kubectl rollout history deployment webcluster
deployment.apps/webcluster
REVISION  CHANGE-CAUSE
1         <none>


[root@k8s-master ~]# kubectl expose deployment webcluster --port 80 --target-port 80 --type NodePort

[root@k8s-master ~]# kubectl get svc
NAME         TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.96.0.1    <none>        443/TCP        24h
webcluster   NodePort    10.99.3.32   <none>        80:30713/TCP   2m54s


[Administrator.DESKTOP-VJ307M3] ➤ curl http://172.25.254.100:30713/hostname.html
webcluster-77c87d9946-2xl44

2.更新业务版本

bash 复制代码 
[root@k8s-master ~]# kubectl set image deployments webcluster myapp=myapp:v2
deployment.apps/webcluster image updated
[root@k8s-master ~]# kubectl annotate deployment webcluster kubernetes.io/change-cause="myappv2" --overwrite			#为此次更新设定标签

[root@k8s-master ~]# kubectl get pods
NAME                          READY   STATUS    RESTARTS   AGE
webcluster-6c8b4bb9d7-hpxf7   1/1     Running   0          7s
webcluster-6c8b4bb9d7-kz46j   1/1     Running   0          8s
[root@k8s-master ~]# kubectl rollout history
daemonset    deployment   statefulset
[root@k8s-master ~]# kubectl rollout history deployment webcluster
deployment.apps/webcluster
REVISION  CHANGE-CAUSE
1         <none>
2         <none>

[2026-08-21 11:30.21]  ~
[Administrator.DESKTOP-VJ307M3] ➤ curl http://172.25.254.100:30713
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

3.版本回退

bash 复制代码 
[root@k8s-master ~]# kubectl rollout history deployment webcluster
deployment.apps/webcluster
REVISION  CHANGE-CAUSE
2         <none>
3         <none>

[root@k8s-master ~]# kubectl rollout undo deployment webcluster --to-revision 3
deployment.apps/webcluster rolled back

[Administrator.DESKTOP-VJ307M3] ➤ curl http://172.25.254.100:30713
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

四.利用yaml文件声明资源

  1. Pod 运行多容器

一个 Pod 可以写多个containers数组，共享网络、存储，但是容器相互隔离。

bash 复制代码 
spec:
  containers:
  - name: nginx
    image: nginx
  - name: busybox
    image: busybox
  1. Pod 暴露端口
bash 复制代码 
spec:
  containers:
  - name: nginx
    image: nginx
    ports:
    - containerPort: 80
  1. Pod 指定环境变量 env
bash 复制代码 
spec:
  containers:
  - name: nginx
    image: nginx
    env:
    - name: "TEST_ENV"
      value: "hello"
  1. 选择运行节点 nodeSelector
bash 复制代码 
spec:
  nodeSelector:
    disk: ssd   # 节点需要打对应标签
  1. 共享宿主机网络 hostNetwork: true
bash 复制代码 
spec:
  hostNetwork: true
  containers:
  - name: nginx
    image: nginx
  1. QoS 资源优先级（3 种）
QoS 类别 条件 优先级
Guaranteed 每个容器 requests.cpu/memory = limits.cpu/memory 最高，OOM 最后被杀
Burstable 设置了 requests，requests ≠ limits 中等
BestEffort 完全不写 requests、limits 资源配置 最低，内存不足最先被 OOM 杀死

  1. 容器重启策略 restartPolicy

  2. Always：无论容器正常退出 / 异常崩溃，永远重启（Deployment 默认）

  3. OnFailure：只有异常退出（返回码≠0）才重启；正常退出不重启

  4. Never ：不管成功失败，绝不重启（一次性任务，Job 使用）

五.pod的生命周期

1.init容器

bash 复制代码 
[root@k8s-master pod]# kubectl run  webserver --image myapp:v1 --dry-run=client -o yaml > init-example.yml
[root@k8s-master pod]# vim init-example.yml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  initContainers:
  - name: busybox
    image: busybox:latest
    command:
    - /bin/sh
    - -c
    - "until test -e /testfile;do echo wating for myservice; sleep 2;done"
  containers:
  - image: myapp:v1
    name: webserver
  restartPolicy: Always


[root@k8s-master pod]# watch -n 1 kubectl get pods  -o wide

[root@k8s-master pod]# kubectl apply -f init-example.yml
pod/webserver created

[root@k8s-master pod]# kubectl exec -it pods/webserver -c busybox -- /bin/sh
/ #
/ # touch /testfile

Livness存活探针

bash 复制代码 
[root@k8s-master pod]# kubectl run  testpod --image  myapp:v1  --dry-run=client -o yaml > livness-example.yaml
[root@k8s-master pod]# vim livness-example.yaml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  containers:
  - image: myapp:v1
    name: webserver
    command: ["/bin/sh", "-c"]
    args:
    - |
      nginx -g "daemon off;"
      sleep 10000
  restartPolicy: Always

[root@k8s-master pod]# kubectl apply -f livness-example.yaml

#监控程序
[root@k8s-master pod]# watch -n 1 kubectl get pods  -o wide

#测试操作：
[root@k8s-master pod]# kubectl exec  -it  pods/webserver -c webserver  -- /bin/sh
# nginx -s stop

#查看pod的状态仍然是runing，但是访问此pod时访问失败
[root@k8s-node2 ~]# curl 10.244.5.47
curl: (7) Failed to connect to 10.244.5.47 port 80: 拒绝连接

2.存活探针livness

bash 复制代码 
[root@k8s-master pod]# kubectl delete -f livness-example.yaml  --force
kind: Pod
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  containers:
  - image: myapp:v1
    name: testpod
    command: ["/bin/sh", "-c"]
    args:
    - |
      nginx -g "daemon off;"
      sleep 10000
    livenessProbe:
      tcpSocket:
        port: 80
      initialDelaySeconds: 3
      periodSeconds: 1
      timeoutSeconds: 1
  restartPolicy: Always


[root@k8s-master pod]# kubectl get pods  -o wide  -w

[root@k8s-master pod]#  kubectl exec -it pods/webserver -c webserver -- /bin/sh
/ #  nginx -s stop
2026/08/23 03:59:45 [notice] 15#15: signal process started

/ # exit
[root@k8s-master pod]# curl 10.244.5.66
curl: (7) Failed to connect to 10.244.5.66 port 80: 拒绝连接
[root@k8s-master pod]# curl 10.244.5.66
curl: (7) Failed to connect to 10.244.5.66 port 80: 拒绝连接
[root@k8s-master pod]# curl 10.244.5.66
curl: (7) Failed to connect to 10.244.5.66 port 80: 拒绝连接
[root@k8s-master pod]# curl 10.244.5.66
curl: (7) Failed to connect to 10.244.5.66 port 80: 拒绝连接
[root@k8s-master pod]# curl 10.244.5.66
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

3.readness 就绪探针

没有就绪探针情况

bash 复制代码 
[root@k8s-master pod]# vim readness-example.yml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  containers:
  - image: myapp:v1
    name: webserver
  restartPolicy: Always

---
apiVersion: v1
kind: Service
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    run: webserver

[root@k8s-master pod]# kubectl apply -f readness-example.yml
pod/webserver unchanged
service/webserver created
[root@k8s-master pod]# kubectl describe  svc webserver
Name:                     webserver
Namespace:                default
Labels:                   run=webserver
Annotations:              <none>
Selector:                 run=webserver
Type:                     ClusterIP
IP Family Policy:         SingleStack
IP Families:              IPv4
IP:                       10.102.162.217
IPs:                      10.102.162.217
Port:                     <unset>  80/TCP
TargetPort:               80/TCP
Endpoints:                10.244.5.67:80
Session Affinity:         None
Internal Traffic Policy:  Cluster
Events:                   <none>

#删除默认发布文件
[root@k8s-master ~]# kubectl exec -it pods/webserver -c webserver -- /bin/sh
/ # cd /usr/share/nginx/
/usr/share/nginx # ls
html
/usr/share/nginx # cd html/
/usr/share/nginx/html # ls
50x.html    index.html
/usr/share/nginx/html # rm -fr index.html
/usr/share/nginx/html # ls
50x.html
/usr/share/nginx/html #

#验证是否在访问service时endpoints中被下架
[root@k8s-master pod]# kubectl describe  svc webserver
Name:                     webserver
Namespace:                default
Labels:                   run=webserver
Annotations:              <none>
Selector:                 run=webserver
Type:                     ClusterIP
IP Family Policy:         SingleStack
IP Families:              IPv4
IP:                       10.102.162.217
IPs:                      10.102.162.217
Port:                     <unset>  80/TCP
TargetPort:               80/TCP
Endpoints:                10.244.5.67:80			#还在
Session Affinity:         None
Internal Traffic Policy:  Cluster
Events:                   <none>


#业务问题
[root@k8s-master pod]# curl  10.102.162.217
<html>
<head><title>403 Forbidden</title></head>
<body bgcolor="white">
<center><h1>403 Forbidden</h1></center>
<hr><center>nginx/1.12.2</center>
</body>
</html>

有就绪探针情况

bash 复制代码 
[root@k8s-master pod]# vim readness-example.yml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  containers:
  - image: myapp:v1
    name: webserver
    readinessProbe:
      httpGet:
        path: /index.html
        port: 80
      initialDelaySeconds: 3
      periodSeconds: 2
      timeoutSeconds: 1
  restartPolicy: Always
---
apiVersion: v1
kind: Service
metadata:
  labels:
    run: webserver
  name: webserver
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    run: webserver


[root@k8s-master pod]# kubectl get pods  -o wide
NAME        READY   STATUS    RESTARTS   AGE   IP            NODE        NOMINATED NODE   READINESS GATES
webserver   1/1     Running   0          7s    10.244.5.69   k8s-node2   <none>           <none>
[root@k8s-master pod]# kubectl describe  svc webserver
Name:                     webserver
Namespace:                default
Labels:                   run=webserver
Annotations:              <none>
Selector:                 run=webserver
Type:                     ClusterIP
IP Family Policy:         SingleStack
IP Families:              IPv4
IP:                       10.110.248.237
IPs:                      10.110.248.237
Port:                     <unset>  80/TCP
TargetPort:               80/TCP
Endpoints:                10.244.5.69:80
Session Affinity:         None
Internal Traffic Policy:  Cluster
Events:                   <none>


#复现问题
/usr/share/nginx/html # command terminated with exit code 137
[root@k8s-master ~]#
/ # rm -fr /usr/share/nginx/html/index.html
/ # echo timinglee > /usr/share/nginx/html/index.html
相关推荐
Yiiz.1 小时前
Kubernetes Pod 与控制器知识点
云原生·容器·kubernetes
qetfw1 小时前
Debian Samba 与 OpenLDAP 统一认证配置
linux·debian
初願致夕霞2 小时前
《为什么公网访问不了内网？正向代理、内网穿透与 NAT 打洞一次说清》
linux·网络·tcp/ip·智能路由器
高磊20052 小时前
Kubernetes Pod 管理实战详解
linux·容器·kubernetes
喜欢吃燃面2 小时前
Linux网络编程深度解析：从Socket API到自定义协议设计
linux·网络·arm开发
流星白龙2 小时前
【Docker】9.Docker 镜像仓库实战
运维·docker·容器
想要成为老金高手2 小时前
Kubernetes 实战笔记（一）：Pod 管理与 kubectl 核心命令全解
笔记·容器·kubernetes
张洛闻Eren2 小时前
云原生k8s【第六课】：K8s 访问控制
运维·docker·云原生·容器·kubernetes·k8s
Kina_C2 小时前
Kubernetes Pod 全生命周期管理：从命令实操到控制器版本更替
云原生·容器·kubernetes
热门推荐
01国内可直接用、免费额度/永久免费的大模型API清单（含 SiliconFlow、火山、阿里、智谱、百度、Kimi、DeepSeek、DMXAPI 等）02GitHub 镜像站点032026 国产 AI 大模型横评：DeepSeek、通义千问、Kimi、文心一言、星火、豆包谁更能打？04微信历史版本含下载地址（ Windows PC | 安卓 | MAC ）及设置微信不更新05Agnes AI 免费 API 接入指南：文本、生图、生视频，一套接口全免费06历年考研数学一、数学二、数学三真题试卷及答案PDF07AI 编程 IDE 全景解析 2026：Agent 全面接管开发链路082026年7月AI圈大地震：GPT-5.6被政府限制、Claude入驻Slack、Anthropic自研芯片09VSCode+CodeX扩展，无法启动10如何新建文件夹？ 电脑新建文件夹的4种方法