云原生(Kubernetes Pod)

匀泪2026-04-12 8:13

实验简介

实验核心目标

  1. 掌握 Pod 两种核心创建方式（命令式、YAML 声明式）的操作流程与差异；
  2. 理解 K8s 核心资源类型（Node、Namespace）与 Pod 的关联关系；
  3. 熟练使用 kubectl 命令完成 Pod/Deployment 的增删改查、扩缩容、版本回滚等运维操作；
  4. 掌握 Pod 多容器部署、网络配置、节点调度等高级配置；
  5. 理解 Pod 生命周期管控机制（Init 容器、存活探针、就绪探针）的作用与配置方式。

实验主要内容模块

1. Pod 资源创建方式

  • 命令式创建 ：直接通过 kubectl run 命令快速创建、查看、删除 Pod，熟悉 Pod 基础状态（Running/ErrImagePull 等）的识别；
  • YAML 声明式创建 ：通过 kubectl create deployment --dry-run 生成模板 YAML，对比 kubectl create（仅创建）与 kubectl apply（创建 / 更新）的差异，掌握 Deployment 控制器管理 Pod 的方式。

2. K8s 核心资源与 kubectl 命令实操

  • 查看 Node 节点状态、生成节点加入命令，理解集群节点架构；
  • 管理 Namespace：创建命名空间、在指定命名空间部署 Pod，实现资源隔离；
  • 核心 kubectl 命令：edit/patch 修改资源、expose 暴露服务、logs 查看日志、exec/attach 进入容器、scale 扩缩容、label 管理标签等。

3. Pod 高级配置与应用部署

  • 单 / 多容器 Pod 部署：验证 Pod 内容器网络互通（共享网络命名空间）；
  • 网络配置：容器端口映射（containerPort/hostPort）、共享宿主机网络（hostNetwork）；
  • 节点调度：通过 nodeSelector 指定 Pod 运行节点；
  • 版本管理：通过 kubectl set image 更新容器镜像，rollout history/undo 实现版本回滚。

4. Pod 生命周期管控

  • Init 容器：配置初始化容器，验证 Init 容器完成后业务容器才启动的逻辑；
  • 存活探针（LivenessProbe）：监控容器健康状态，容器异常时自动重启；
  • 就绪探针（ReadinessProbe）：验证容器服务就绪状态，未就绪时剔除 Service 端点，避免流量分发至未就绪 Pod。

实验价值

通过本实验可掌握 K8s 中 Pod 从创建到销毁的全生命周期管理，理解声明式 API 的核心优势，掌握生产环境中 Pod 运维的核心命令与配置技巧，为后续 K8s 服务编排、高可用部署打下基础。

1. 基础 Pod 与 Deployment 操作 (建立、扩缩容、暴露服务)

复制代码 
# 工作区准备
[root@master ~]# mkdir -p ~/pod-labs && cd ~/pod-labs

# 命令式：直接运行并查看
[root@master pod-labs]# kubectl run webpod --image nginx:latest --port 80
pod/webpod created
[root@master ~]# kubectl get pods -o wide
NAME     READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
webpod   1/1     Running   0          6s    10.244.2.3   node2   <none>           <none>
[root@master ~]# kubectl delete pod webpod
pod "webpod" deleted from default namespace

# 声明式：生成 Deployment 的 YAML 并应用
[root@master ~]# kubectl create deployment test --image nginx --replicas 1 --dry-run=client -o yaml > test.yml
[root@master ~]# kubectl apply -f test.yml
deployment.apps/test created

# 扩缩容测试
[root@master ~]# kubectl scale deployment test --replicas 3
deployment.apps/test scaled
[root@master ~]# kubectl get pods
NAME                    READY   STATUS    RESTARTS   AGE
test-56848fd9dc-bgx58   1/1     Running   0          4s
test-56848fd9dc-gcq8m   1/1     Running   0          18s
test-56848fd9dc-q9b4h   1/1     Running   0          4s
[root@master ~]# kubectl scale deployment test --replicas 1
deployment.apps/test scaled
[root@master ~]# kubectl get pods
NAME                    READY   STATUS    RESTARTS   AGE
test-56848fd9dc-bgx58   1/1     Running   0          12s

# 暴露服务并通过动态提取 ClusterIP 进行访问测试
[root@master ~]# kubectl expose deployment test --port 80 --target-port 80
service/test exposed
[root@master ~]# SVC_IP=$(kubectl get svc test -o jsonpath='{.spec.clusterIP}')
[root@master ~]# curl $SVC_IP
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
html { color-scheme: light dark; }
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.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>

<p><em>Thank you for using nginx.</em></p>
</body>
</html>

# 清理当前实验环境
[root@master ~]# kubectl delete -f test.yml
deployment.apps "test" deleted from default namespace
[root@master ~]# kubectl delete svc test
service "test" deleted from default namespace

2. 命名空间与标签管理

复制代码 
# 创建并使用独立的 namespace
[root@master ~]# kubectl create namespace timinglee
namespace/timinglee created
[root@master ~]# kubectl -n timinglee run testpod --image nginx:latest
pod/testpod created

# 验证
[root@master ~]# kubectl -n timinglee get pods
NAME      READY   STATUS    RESTARTS   AGE
testpod   1/1     Running   0          11s

# 为 Pod 打标签与验证
[root@master ~]# kubectl label pods testpod -n timinglee name=lee
pod/testpod labeled
[root@master ~]# kubectl get pods -n timinglee --show-labels
NAME      READY   STATUS    RESTARTS   AGE   LABELS
testpod   1/1     Running   0          26s   name=lee,run=testpod

# 删除指定标签 (注意最后的减号)
[root@master ~]# kubectl label pods testpod -n timinglee name-
pod/testpod unlabeled

# 统一清理
[root@master ~]# kubectl delete namespace timinglee
namespace "timinglee" deleted

3. 应用版本滚动更新与回滚

复制代码 
# 部署 v2 版本并暴露服务
[root@master ~]# kubectl create deployment webcluster --image myapp:v2 --replicas 2
deployment.apps/webcluster created
[root@master ~]# kubectl expose deployment webcluster --port 80 --target-port 80
service/webcluster exposed

# 获取 IP 并验证当前版本
[root@master ~]# docker load -i myapp.tar.gz
d39d92664027: Loading layer  4.232MB/4.232MB
8460a579ab63: Loading layer  11.61MB/11.61MB
c1dc81a64903: Loading layer  3.584kB/3.584kB
68695a6cfd7d: Loading layer  4.608kB/4.608kB
05a9e65e2d53: Loading layer  16.38kB/16.38kB
a0d2c4392b06: Loading layer   7.68kB/7.68kB
Loaded image: timinglee/myapp:v1
Loaded image: timinglee/myapp:v2
[root@master ~]# docker tag timinglee/myapp:v1 reg.timinglee.org/library/myapp:v1
[root@master ~]# docker tag timinglee/myapp:v2 reg.timinglee.org/library/myapp:v2
[root@master ~]# docker push reg.timinglee.org/library/myapp:v1
The push refers to repository [reg.timinglee.org/library/myapp]
a0d2c4392b06: Pushed 
05a9e65e2d53: Pushed 
68695a6cfd7d: Pushed 
c1dc81a64903: Pushed 
8460a579ab63: Pushed 
d39d92664027: Pushed 
v1: digest: sha256:9eeca44ba2d410e54fccc54cbe9c021802aa8b9836a0bcf3d3229354e4c8870e size: 1569
[root@master ~]# docker push reg.timinglee.org/library/myapp:v2
The push refers to repository [reg.timinglee.org/library/myapp]
05a9e65e2d53: Layer already exists 
68695a6cfd7d: Layer already exists 
c1dc81a64903: Layer already exists 
8460a579ab63: Layer already exists 
d39d92664027: Layer already exists 
v2: digest: sha256:5f4afc8302ade316fc47c99ee1d41f8ba94dbe7e3e7747dd87215a15429b9102 size: 1362
[root@master ~]# kubectl set image deployments webcluster myapp=reg.timinglee.org/library/myapp:v2
deployment.apps/webcluster image updated
[root@master ~]# kubectl get pods
NAME                          READY   STATUS    RESTARTS   AGE
webcluster-6fbfd55778-qf96n   1/1     Running   0          4s
webcluster-6fbfd55778-v84t9   1/1     Running   0          6s
[root@master ~]# SVC_IP=$(kubectl get svc webcluster -o jsonpath='{.spec.clusterIP}')
[root@master ~]# curl $SVC_IP
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

# 更新版本至 v1
[root@master ~]# kubectl set image deployments webcluster myapp=myapp:v1
deployment.apps/webcluster image updated
[root@master ~]# kubectl rollout status deployment webcluster
deployment "webcluster" successfully rolled out
[root@master ~]# curl $SVC_IP
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

# 一键回滚至上一个版本 (v2)
[root@master ~]# kubectl rollout undo deployment webcluster --to-revision 1
deployment.apps/webcluster rolled back
[root@master ~]# kubectl rollout status deployment webcluster
deployment "webcluster" successfully rolled out
[root@master ~]# curl $SVC_IP
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

# 清理
[root@master ~]# kubectl delete deployment webcluster
deployment.apps "webcluster" deleted from default namespace
[root@master ~]# kubectl delete svc webcluster
service "webcluster" deleted from default namespace

4. Pod 进阶配置 (节点选择、端口映射、共享网络)

复制代码 
# 场景 A: 指定部署在 Node1 且映射宿主机端口
[root@master ~]# cat <<EOF > 5test.yml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: lee1
  name: lee1
spec:
  nodeSelector:
    kubernetes.io/hostname: node1
  containers:
  - image: myapp:v1
    name: myappv1
    ports:
    - name: webport
      containerPort: 80
      hostPort: 80
      protocol: TCP
EOF
[root@master ~]# kubectl apply -f 5test.yml
pod/lee1 created
[root@master ~]# kubectl get pods -o wide
NAME   READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
lee1   1/1     Running   0          3s    10.244.1.10   node1   <none>           <none>

# 清理
[root@master ~]# kubectl delete -f 5test.yml
pod "lee1" deleted from default namespace

# 场景 B: 共享 Node1 宿主机网络空间
[root@master ~]# cat <<EOF > 6test.yml
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: lee1
  name: lee1
spec:
  hostNetwork: true
  nodeSelector:
    kubernetes.io/hostname: node1
  containers:
  - image: busybox:latest
    name: busybox
    command:
      - /bin/sh
      - -c
      - sleep 1000
EOF
[root@master ~]# kubectl apply -f 6test.yml
pod/lee1 created
[root@master ~]# sleep 3

# 此时查看到的网卡信息为 node1 的真实物理网卡信息
[root@master ~]# kubectl exec -it lee1 -c busybox -- ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq qlen 1000
    link/ether 00:0c:29:f2:ef:6c brd ff:ff:ff:ff:ff:ff
    inet 172.25.254.10/24 brd 172.25.254.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::fc4:f2e3:a2e:9385/64 scope link 
       valid_lft forever preferred_lft forever
3: docker0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue 
    link/ether e6:4b:a0:cb:46:96 brd ff:ff:ff:ff:ff:ff
    inet 172.17.0.1/16 brd 172.17.255.255 scope global docker0
       valid_lft forever preferred_lft forever
4: flannel.1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue 
    link/ether 06:8a:d6:36:45:2d brd ff:ff:ff:ff:ff:ff
    inet 10.244.1.0/32 scope global flannel.1
       valid_lft forever preferred_lft forever
    inet6 fe80::48a:d6ff:fe36:452d/64 scope link 
       valid_lft forever preferred_lft forever
5: cni0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue qlen 1000
    link/ether 62:c8:67:ba:53:c4 brd ff:ff:ff:ff:ff:ff
    inet 10.244.1.1/24 brd 10.244.1.255 scope global cni0
       valid_lft forever preferred_lft forever
    inet6 fe80::60c8:67ff:feba:53c4/64 scope link 
       valid_lft forever preferred_lft forever

# 清理
kubectl delete -f 6test.yml

5. 生命周期管理 (Init 容器与就绪探针)

复制代码 
# 场景 A: Init 容器控制主容器启动
[root@master ~]# cat <<EOF > init.yml
apiVersion: v1
kind: Pod
metadata:
  name: lee-init
spec:
  initContainers:
  - name: init-myservice
    image: busybox
    command: ["sh","-c","until test -e /testfile;do echo wating; sleep 2;done"]
  containers:
  - image: myapp:v1
    name: myappv1
EOF
[root@master ~]# kubectl apply -f init.yml
pod/lee-init created

# 此时 Pod 状态会卡在 Init:0/1
[root@master ~]# kubectl get pods lee-init
NAME       READY   STATUS     RESTARTS   AGE
lee-init   0/1     Init:0/1   0          13s

# 通过命令强行给 Init 容器传入启动信号 (/testfile)
[root@master ~]# kubectl exec -it lee-init -c init-myservice -- touch /testfile
[root@master ~]# sleep 3

# 此时主容器顺利启动，状态变为 Running
[root@master ~]# kubectl get pods lee-init
NAME       READY   STATUS    RESTARTS   AGE
lee-init   1/1     Running   0          36s
[root@master ~]# kubectl delete -f init.yml
pod "lee-init" deleted from default namespace

# 场景 B: 就绪探针 (Readiness Probe)
[root@master ~]# cat <<EOF > probe.yml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: webcluster-probe
spec:
  replicas: 1
  selector:
    matchLabels:
      app: webcluster-probe
  template:
    metadata:
      labels:
        app: webcluster-probe
    spec:
      containers:
      - image: myapp:v1
        name: myapp
        readinessProbe:
          httpGet:
            path: /test.html
            port: 80
          initialDelaySeconds: 1
          periodSeconds: 3
          timeoutSeconds: 1
---
apiVersion: v1
kind: Service
metadata:
  name: webcluster-probe
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    app: webcluster-probe
EOF
[root@master ~]# kubectl apply -f probe.yml
deployment.apps/webcluster-probe created
service/webcluster-probe created
[root@master ~]# sleep 3

# 因为没有 test.html，容器状态为 Running，但 READY 为 0/1 (流量不接入)
[root@master ~]# kubectl get pods -l app=webcluster-probe
NAME                                READY   STATUS    RESTARTS   AGE
webcluster-probe-7ccc496879-vtc5k   0/1     Running   0          19s

# 动态获取 Pod 名称，并在容器内创建所需的探针验证文件
[root@master ~]# POD_NAME=$(kubectl get pods -l app=webcluster-probe -o jsonpath="{.items[0].metadata.name}")
[root@master ~]# kubectl exec -it $POD_NAME -c myapp -- sh -c "echo timinglee > /usr/share/nginx/html/test.html"

# 等待3秒探针重试后，READY 会变为 1/1
[root@master ~]# sleep 3
[root@master ~]# kubectl get pods -l app=webcluster-probe
NAME                                READY   STATUS    RESTARTS   AGE
webcluster-probe-7ccc496879-vtc5k   1/1     Running   0          52s

# 测试销毁探针文件，READY 会重新掉回 0/1
[root@master ~]# kubectl exec -it $POD_NAME -c myapp -- rm -fr /usr/share/nginx/html/test.html
[root@master ~]# sleep 3
[root@master ~]# kubectl get pods -l app=webcluster-probe
NAME                                READY   STATUS    RESTARTS   AGE
webcluster-probe-7ccc496879-vtc5k   0/1     Running   0          81s

# 最终清理
[root@master ~]# kubectl delete -f probe.yml
deployment.apps "webcluster-probe" deleted from default namespace
service "webcluster-probe" deleted from default namespace
相关推荐
砍材农夫2 小时前
番外篇-ollama一步一步docker安装部署
运维·docker·容器
江湖有缘2 小时前
Docker部署EnBizCard教程：从零搭建专属开源数字名片平台
docker·容器·开源
架构师老Y2 小时前
009、容器编排实战：Kubernetes上的Python服务
python·容器·kubernetes
问简11 小时前
docker 镜像相关
运维·docker·容器
Benszen12 小时前
Docker容器化技术实战指南
运维·docker·容器
lin_dec+12 小时前
Serverless：零成本按需计算的未来
云原生·serverless
Hommy8813 小时前
【开源剪映小助手】Docker 部署
docker·容器·开源·github·aigc
斯普信云原生组14 小时前
Prometheus 环境监控虚机 Redis 方案（生产实操版）
运维·docker·容器
喵了几个咪14 小时前
如何在 Superset Docker 容器中安装 MySQL 驱动
mysql·docker·容器·superset
热门推荐
01GitHub 镜像站点02一周AI热点速览（2026.03.31-04.06）：GPT-6曝光、谷歌开源Gemma 4、资本狂飙与模型军备竞赛03OpenClaw 请求超时 llm request timed out 怎么解决？3 种方案实测，附完整排查流程04基于 Docker 部署 Hermes Agent 并接入飞书机器人的完整指南05VMware Workstation Pro 17 虚拟机完整安装教程（2026最新）06Oh My Codex 快速使用指南07AI Weekly | 2026年4月第二周 · GitHub热门项目与AI发展趋势深度解析08CodeBuddy与WorkBuddy深度对比：腾讯两款AI工具差异及实操指南09AI 编程效率翻倍：Superpowers Skills 上手清单 + 完整指南10实测！Gemma 4 成功跑在安卓手机上：离线 AI 助手终于来了