6.k8s中Jobs-CronJobs-configmap-备份etcd

认真的薛薛2026-02-19 21:28

K8S常见的控制器之Jobs和CronJobs

Jobs控制器

概念

所谓的Jobs就是K8S集群用于处理一次性任务的控制器。

编写资源清单

示例：简单打印一个句子

yaml 复制代码 
# 01-jobs-pi.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: jobs-py
spec:
  # 定义Pod的模板
  template:
    spec:
      containers:
      - name: py
        image: harbor250.oldboyedu.com/oldboyedu-casedemo/python:3.9.16-alpine3.16
        command:
        - python3 
        - -c 
        - "print('https://www.oldboyedu.com/')"
      restartPolicy: Never
 #never不重启
  # 如果容器执行失败，则重试的次数，重试会重新创建新的Pod
  backoffLimit: 4

查看日志

bash 复制代码 
[root@master231 jobs]# kubectl logs -f jobs-py-rj228 
https://www.oldboyedu.com/

Pod的五个阶段

  1. Pending（悬决）

    • Pod已被Kubernetes系统接受，但有一个或者多个容器尚未创建亦未运行
    • 此阶段包括等待Pod被调度的时间和通过网络下载镜像的时间

  2. Running（运行中）

    • Pod已经绑定到了某个节点，Pod中所有的容器都已被创建
    • 至少有一个容器仍在运行，或者正处于启动或重启状态

  3. Succeeded（成功）

    • Pod中的所有容器都已成功终止，并且不会再重启

  4. Failed（失败）

    • Pod中的所有容器都已终止，并且至少有一个容器是因为失败终止
    • 也就是说，容器以非0状态退出或者被系统终止，且未被设置为自动重启

  5. Unknown（未知）

    • 因为某些原因无法取得Pod的状态
    • 这种情况通常是因为与Pod所在主机通信失败

K8S常见的控制器之CronJobs

概念

全称为"cronjobs"，其并不直接管理Pod，而是底层周期性调用Job控制器。

编写资源清单

yaml 复制代码 
# 01-cj-demo.yaml
apiVersion: batch/v1
kind: CronJob
metadata:
  name: hello
spec:
  # 指定调度的周期，分别表示: 分，时，日，月，周
  schedule: "* * * * *"
  # 定义Job控制器的模板
  jobTemplate:
    spec:
      template:
        spec:
          volumes:
          - name: dt
            hostPath:
              path: /etc/localtime
          containers:
          - name: hello
            image: harbor250.oldboyedu.com/oldboyedu-casedemo/busybox:1.28
            volumeMounts:
            - name: dt
              mountPath: /etc/localtime
            command:
            - /bin/sh
            - -c
            - date -R; echo Hello from the Kubernetes cluster
          restartPolicy: OnFailure

测试验证

bash 复制代码 
[root@master231 cronjobs]# kubectl apply -f  01-cj-demo.yaml
cronjob.batch/hello created

# 测试验证 cronjobs 创建job ，job创建po：会每分钟同步一次时间然后输出一次
[root@master231 cronjobs]# kubectl get cj,job,po
NAME                  SCHEDULE    SUSPEND   ACTIVE   LAST SCHEDULE   AGE
cronjob.batch/hello   * * * * *   False     0        7s              93s

NAME                       COMPLETIONS   DURATION   AGE
job.batch/hello-29312735   1/1           2s         67s
job.batch/hello-29312736   1/1           3s         7s

NAME                       READY   STATUS      RESTARTS   AGE
pod/hello-29312735-qq642   0/1     Completed   0          67s
pod/hello-29312736-gt2lt   0/1     Completed   0          7s

[root@master231 cronjobs]# kubectl logs hello-29312735-qq642 
Thu, 25 Sep 2025 09:35:00 +0800
Hello from the Kubernetes cluster

[root@master231 cronjobs]# kubectl logs hello-29312736-gt2lt  # 很明显，时间相差1min
Thu, 25 Sep 2025 09:36:00 +0800
Hello from the Kubernetes cluster

重点掌握

Pod长期处于Terminating的解决方案：基于etcd删除Pod

1. 部署测试服务

bash 复制代码 
[root@master231 pods]# kubectl get nodes 
NAME        STATUS   ROLES                  AGE     VERSION
master231   Ready    control-plane,master   5d23h   v1.23.17
worker232   Ready    <none>                 5d23h   v1.23.17
worker233   Ready    <none>                 5d23h   v1.23.17

[root@master231 pods]# cat 13-pods-nodeName.yaml
apiVersion: v1
kind: Pod
metadata:
  name: xixi
spec:
  nodeName: worker233
  containers:
  - name: c1
    image: registry.cn-hangzhou.aliyuncs.com/yinzhengjie-k8s/apps:v2

[root@master231 pods]# kubectl apply -f  13-pods-nodeName.yaml
pod/xixi created

[root@master231 pods]# kubectl get pods -o wide
NAME   READY   STATUS    RESTARTS   AGE   IP             NODE        NOMINATED NODE   READINESS GATES
xixi   1/1     Running   0          4s    10.100.2.185   worker233   <none>           <none>

2. 将指定节点关机模拟故障

bash 复制代码 
[root@worker233 ~]# init 0

3. 再次查看node和pod状态

bash 复制代码 
[root@master231 pods]# kubectl get nodes 
NAME        STATUS     ROLES                  AGE     VERSION
master231   Ready      control-plane,master   5d23h   v1.23.17
worker232   Ready      <none>                 5d23h   v1.23.17
worker233   NotReady   <none>                 5d23h   v1.23.17

[root@master231 pods]# kubectl get pods -o wide
NAME   READY   STATUS    RESTARTS   AGE   IP             NODE        NOMINATED NODE   READINESS GATES
xixi   1/1     Running   0          93s   10.100.2.185   worker233   <none>           <none>

[root@master231 pods]# kubectl describe pod xixi 
# 最后一行  Warning  NodeNotReady  2m4s   node-controller  Node is not ready

[root@master231 pods]# curl  --connect-timeout 2 10.100.2.185  # 不难发现，此时Pod无法提供正常服务
curl: (28) Connection timeout after 2001 ms

[root@master231 pods]# kubectl delete pod xixi  # 我们会发现无法删除该Pod
pod "xixi" deleted一直卡着

4. 检查Pod状态

bash 复制代码 
[root@master231 ~]# kubectl get pods -o wide
NAME   READY   STATUS        RESTARTS   AGE     IP             NODE        NOMINATED NODE   READINESS GATES
xixi   1/1     Terminating   0          8m26s   10.100.2.185   worker233   <none>           <none>

[root@master231 ~]# kubectl describe pod xixi 
Name:                      xixi
Namespace:                 default
Priority:                  0
Node:                      worker233/10.0.0.233
Start Time:                Thu, 25 Sep 2025 10:15:59 +0800
Labels:                    <none>
Annotations:               <none>
Status:                    Terminating (lasts 2m4s)  # 发现Pod处于正在终止状态，已经超过垃圾回收时间30s，已经耗时2m4s
Termination Grace Period:  30s  # 30s后强行退出。
IP:                        10.100.2.185
...
# 怎么办呢，数据存在etcd所以访问etcd

5. 连接etcd数据库删除数据

bash 复制代码 
# 5.1 下载etcd的软件包
wget https://github.com/etcd-io/etcd/releases/download/v3.5.22/etcd-v3.5.22-linux-amd64.tar.gz

# 5.2 解压etcd的二进制程序包到PATH环境变量路径
[root@master231 ~]# tar -xf etcd-v3.5.22-linux-amd64.tar.gz -C /usr/local/bin etcd-v3.5.22-linux-amd64/etcdctl  --strip-components=1 

[root@master231 ~]# ll /usr/local/bin/etcd*
-rwxr-xr-x 1 yinzhengjie yinzhengjie 19062936 Jul 23 02:20 /usr/local/bin/etcdctl*

[root@master231 ~]# etcdctl version
etcdctl version: 3.5.22
API version: 3.5

# 5.3 检查etcd所在的worker节点
[root@master231 ~]# kubectl get pods -n kube-system -l component=etcd -o wide
NAME             READY   STATUS    RESTARTS      AGE     IP           NODE        NOMINATED NODE   READINESS GATES
etcd-master231   1/1     Running   9 (81m ago)   5d23h   10.0.0.231   master231   <none>           <none>

[root@master231 ~]# ss -ntl | grep 2379
LISTEN 0      4096       127.0.0.1:2379       0.0.0.0:*          
LISTEN 0      4096      10.0.0.231:2379       0.0.0.0:*          

# 5.4 连接etcd集群：etcd后面指定的是认证文件和节点信息
[root@master231 ~]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key  endpoint status --write-out=table
+-------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
|        ENDPOINT         |        ID        | VERSION | DB SIZE | IS LEADER | IS LEARNER | RAFT TERM | RAFT INDEX | RAFT APPLIED INDEX | ERRORS |
+-------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
| https://10.0.0.231:2379 | 3b1681ce69008ae3 |   3.5.6 |  3.9 MB |      true |      false |        11 |     222010 |             222010 |        |
+-------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+

[root@master231 ~]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key  get "" --prefix --keys-only | grep -v "^$" | wc -l
675

# 5.5 删除etcd数据验证pod信息是否丢失
[root@master231 ~]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key  get "" --prefix --keys-only | grep pods
/registry/pods/default/xixi
/registry/pods/kube-flannel/kube-flannel-ds-7fpf5
/registry/pods/kube-flannel/kube-flannel-ds-kxcgq
/registry/pods/kube-flannel/kube-flannel-ds-vzx4x
/registry/pods/kube-system/coredns-6d8c4cb4d-s85bq
/registry/pods/kube-system/coredns-6d8c4cb4d-tdrnj
/registry/pods/kube-system/etcd-master231
/registry/pods/kube-system/kube-apiserver-master231
/registry/pods/kube-system/kube-controller-manager-master231
/registry/pods/kube-system/kube-proxy-55jhz
/registry/pods/kube-system/kube-proxy-7ffbs
/registry/pods/kube-system/kube-proxy-xgqln
/registry/pods/kube-system/kube-scheduler-master231
/registry/pods/metallb-system/controller-644c958987-9xbtc
/registry/pods/metallb-system/speaker-hfb5k
/registry/pods/metallb-system/speaker-p27z5
/registry/pods/metallb-system/speaker-vtxk8

[root@master231 ~]# kubectl get pods -o wide
NAME   READY   STATUS        RESTARTS   AGE   IP             NODE        NOMINATED NODE   READINESS GATES
xixi   1/1     Terminating   0          20m   10.100.2.185   worker233   <none>           <none>

# 删除这个报错的节点
[root@master231 ~]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key  del /registry/pods/default/xixi
1

[root@master231 ~]# kubectl get pods -o wide
No resources found in default namespace.
# 所以可以直接操作etcd，删除pod

课堂练习之周期性备份etcd数据库

1. 准备镜像

bash 复制代码 
[root@master231 ~]# cp /usr/local/bin/etcdctl ./

[root@master231 ~]# cat Dockerfile 
FROM registry.cn-hangzhou.aliyuncs.com/yinzhengjie-k8s/apps:v1

LABEL school=oldboyedu \
      class=linux99 \
      auther=JasonYin

COPY etcdctl /usr/local/bin

CMD ["tail","-f","/etc/hosts"]

[root@master231 ~]# docker build -t  harbor250.oldboyedu.com/oldboyedu-casedemo/etcdctl:v1 .
[root@master231 ~]# docker push  harbor250.oldboyedu.com/oldboyedu-casedemo/etcdctl:v1 

# 温馨提示:不要使用官方的镜像，有坑，很多命令都没有，比如tail,sleep,date都没有！！！

2. 准备etcd的证书文件

bash 复制代码 
[root@master231 ~]# mkdir -p /yinzhengjie/data/nfs-server/casedemo/etcd/certs
[root@master231 ~]# cp /etc/kubernetes/pki/etcd/{ca.crt,server.crt,server.key} /yinzhengjie/data/nfs-server/casedemo/etcd/certs

[root@master231 ~]# ll /yinzhengjie/data/nfs-server/casedemo/etcd/certs
total 20
drwxr-xr-x 2 root root 4096 Sep 25 10:42 ./
drwxr-xr-x 3 root root 4096 Sep 25 10:41 ../
-rw-r--r-- 1 root root 1086 Sep 25 10:42 ca.crt
-rw-r--r-- 1 root root 1200 Sep 25 10:42 server.crt
-rw------- 1 root root 1675 Sep 25 10:42 server.key

3. 准备备份的数据目录

bash 复制代码 
[root@master231 ~]# mkdir /yinzhengjie/data/nfs-server/casedemo/etcd/backup

4. 编写资源清单

yaml 复制代码 
# 03-cj-backup-etcd.yaml
apiVersion: batch/v1
kind: CronJob
metadata:
  name: cj-backup-etcd
spec:
  schedule: "* * * * *"
  jobTemplate:
    spec:
      template:
        spec:
          # 就是将宿主机里的数据，挂载到容器里面，分别是时间，认证文件，快照数据，共享完直接在容器里面执行快照的命令
          volumes:   
          - name: dt
            hostPath:
              path: /etc/localtime

          - name: certs
            nfs:
              server: 10.0.0.231
              path: /yinzhengjie/data/nfs-server/casedemo/etcd/certs

          - name: data
            nfs:
              server: 10.0.0.231
              path: /yinzhengjie/data/nfs-server/casedemo/etcd/backup

          containers:
          - name: hello
            image: harbor250.oldboyedu.com/oldboyedu-casedemo/etcdctl:v1

            volumeMounts:
            - name: dt
              mountPath: /etc/localtime

            - name: certs
              mountPath: /etc/kubernetes/pki/etcd

            - name: data
              mountPath: /oldboyedu/data

            command:
            - /bin/sh
            - -c
            - "etcdctl --endpoints='https://10.0.0.231:2379' --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key snapshot save /oldboyedu/data/oldboyedu-etcd-$(date +%F_%T).backup"
          restartPolicy: OnFailure

5. 创建资源

bash 复制代码 
[root@master231 case-demo]# kubectl apply -f  03-cj-backup-etcd.yaml 
cronjob.batch/cj-backup-etcd created

[root@master231 case-demo]# kubectl get cj,job,po -o wide
NAME                           SCHEDULE    SUSPEND   ACTIVE   LAST SCHEDULE   AGE   CONTAINERS   IMAGES                                                  SELECTOR
cronjob.batch/cj-backup-etcd   * * * * *   False     1        2s              22s   hello        harbor250.oldboyedu.com/oldboyedu-casedemo/etcdctl:v1   <none>

NAME                                COMPLETIONS   DURATION   AGE   CONTAINERS   IMAGES                                                  SELECTOR
job.batch/cj-backup-etcd-29312860   0/1           2s         2s    hello        harbor250.oldboyedu.com/oldboyedu-casedemo/etcdctl:v1   controller-uid=cfcbf2df-9906-4cac-b7a3-a8a7cf94a923

NAME                                READY   STATUS      RESTARTS   AGE   IP             NODE        NOMINATED NODE   READINESS GATES
pod/cj-backup-etcd-29312860-nwgvv   0/1     Completed   0          2s    10.100.1.110   worker232   <none>           <none>

6. 验证数据

bash 复制代码 
[root@master231 ~]# ll /yinzhengjie/data/nfs-server/casedemo/etcd/backup/
total 3820
drwxr-xr-x 2 root root    4096 Sep 25 11:43 ./
drwxr-xr-x 4 root root    4096 Sep 25 11:08 ../
-rw------- 1 root root 3899424 Sep 25 11:43 oldboyedu-etcd-2025-09-25_11:43:00.backup

基于备份的数据进行还原K8S集群环境（故障面试题）生产环境一定要备份

1. 模拟故障，删除etcd的所有数据，直接node全没了

bash 复制代码 
[root@master231 ~]# kubectl get nodes 
NAME        STATUS     ROLES                  AGE   VERSION
master231   Ready      control-plane,master   6d    v1.23.17
worker232   Ready      <none>                 6d    v1.23.17
worker233   NotReady   <none>                 6d    v1.23.17

[root@master231 ~]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key  del "" --prefix  # 删除所有的数据

[root@master231 ~]# rm -rf /var/lib/etcd/*

[root@master231 ~]# reboot 

[root@master231 ~]# ss -ntl | grep 2379
LISTEN 0      4096      10.0.0.231:2379       0.0.0.0:*          
LISTEN 0      4096       127.0.0.1:2379       0.0.0.0:*          

[root@master231 ~]# kubectl get nodes -o wide # 很明显数据已经不准确啦！
NAME        STATUS   ROLES    AGE   VERSION    INTERNAL-IP   EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION       CONTAINER-RUNTIME
master231   Ready    <none>   60s   v1.23.17   10.0.0.231    <none>        Ubuntu 22.04.4 LTS   5.15.0-153-generic   docker://20.10.24

2. 停止etcd服务

bash 复制代码 
[root@master231 ~]# cd /etc/kubernetes/manifests/     # 这是静态pod的目录

[root@master231 manifests]# ll
total 24
drwxr-xr-x 2 root root 4096 Sep 24 16:21 ./
drwxr-xr-x 4 root root 4096 Sep 19 11:03 ../
-rw------- 1 root root 2280 Sep 19 11:03 etcd.yaml
-rw------- 1 root root 4068 Sep 24 16:20 kube-apiserver.yaml
-rw------- 1 root root 3546 Sep 19 11:03 kube-controller-manager.yaml
-rw------- 1 root root 1465 Sep 19 11:03 kube-scheduler.yaml

[root@master231 manifests]# mv etcd.yaml /opt/       # 将启动的etcd的pod从静态pod目录移出先

[root@master231 manifests]# ss -ntl | grep 2379  # 此时etcd服务停止

3. 恢复数据

bash 复制代码 
[root@master231 manifests]# mv /var/lib/etcd/ /opt/  # 之前的数据暂时备份，之前的数据移走

[root@master231 manifests]# ll /var/lib/etcd
ls: cannot access '/var/lib/etcd': No such file or directory

# 还原
[root@master231 manifests]# etcdctl --endpoints="https://10.0.0.231:2379" --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key snapshot restore /yinzhengjie/data/nfs-server/casedemo/etcd/backup/oldboyedu-etcd-2025-09-25_11\:43\:00.backup --data-dir=/var/lib/etcd/

Deprecated: Use `etcdutl snapshot restore` instead.

2025-09-25T12:05:30+08:00	info	snapshot/v3_snapshot.go:265	restoring snapshot	{"path": "/yinzhengjie/data/nfs-server/casedemo/etcd/backup/oldboyedu-etcd-2025-09-25_11:43:00.backup", "wal-dir": "/var/lib/etcd/member/wal", "data-dir": "/var/lib/etcd/", "snap-dir": "/var/lib/etcd/member/snap", "initial-memory-map-size": 0}
2025-09-25T12:05:31+08:00	info	membership/store.go:138	Trimming membership information from the backend...
2025-09-25T12:05:31+08:00	info	membership/cluster.go:421	added member	{"cluster-id": "cdf818194e3a8c32", "local-member-id": "0", "added-peer-id": "8e9e05c52164694d", "added-peer-peer-urls": ["http://localhost:2380"], "added-peer-is-learner": false}
2025-09-25T12:05:31+08:00	info	snapshot/v3_snapshot.go:293	restored snapshot	{"path": "/yinzhengjie/data/nfs-server/casedemo/etcd/backup/oldboyedu-etcd-2025-09-25_11:43:00.backup", "wal-dir": "/var/lib/etcd/member/wal", "data-dir": "/var/lib/etcd/", "snap-dir": "/var/lib/etcd/member/snap", "initial-memory-map-size": 0}

[root@master231 manifests]# ll /var/lib/etcd
total 12
drwx------  3 root root 4096 Sep 25 12:05 ./
drwxr-xr-x 65 root root 4096 Sep 25 12:05 ../
drwx------  4 root root 4096 Sep 25 12:05 member/

[root@master231 manifests]# ll /var/lib/etcd/member/
total 16
drwx------ 4 root root 4096 Sep 25 12:05 ./
drwx------ 3 root root 4096 Sep 25 12:05 ../
drwx------ 2 root root 4096 Sep 25 12:05 snap/
drwx------ 2 root root 4096 Sep 25 12:05 wal/

4. 启动etcd服务：静态pod启动

bash 复制代码 
[root@master231 manifests]# mv /opt/etcd.yaml ./

[root@master231 manifests]# ss -ntl | grep 2379
LISTEN 0      4096      10.0.0.231:2379       0.0.0.0:*          
LISTEN 0      4096       127.0.0.1:2379       0.0.0.0:*

5. 验证数据是否恢复

bash 复制代码 
[root@master231 manifests]# kubectl get nodes -o wide
NAME        STATUS     ROLES                  AGE    VERSION    INTERNAL-IP   EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION       CONTAINER-RUNTIME
master231   Ready      control-plane,master   6d1h   v1.23.17   10.0.0.231    <none>        Ubuntu 22.04.4 LTS   5.15.0-153-generic   docker://20.10.24
worker232   Ready      <none>                 6d1h   v1.23.17   10.0.0.232    <none>        Ubuntu 22.04.4 LTS   5.15.0-153-generic   docker://20.10.24
worker233   NotReady   <none>                 6d1h   v1.23.17   10.0.0.233    <none>        Ubuntu 22.04.4 LTS   5.15.0-153-generic   docker://20.10.24

重点掌握：K8S集群资源之ConfigMap

概念

简称"cm"，主要存储的是配置信息，只要是配置相关的信息都可以存储在该资源中。cm资源可以被k8s的pod以环境变量的方式注入，也可以基于存储卷的方式注入。

资源清单

yaml 复制代码 
# 01-cm-conf.yaml
apiVersion: v1
kind: ConfigMap
metadata:
  name: cm-conf
# 一个是key，一个是value
data:
  school: oldboyedu
  class: linux99

# my.cnf是key，下面内容是value，写了个配置文件
  my.cnf: |
    [mysqld]
    datadir=/oldboyedu/data/mysql80
    basedir=/oldboyedu/softwares/mysql80
    socket=/tmp/mysql80
    [client]
    username=linux99
    password=oldboyedu

# 写了个脚本
  start.sh: |
    #!/bin/sh
    mkdir /tmp/yinzhengjie
    mount -t tmpfs -o size=90M tmpfs /tmp/yinzhengjie/
    dd if=/dev/zero of=/tmp/yinzhengjie/bigfile.log
    sleep 30
    rm /tmp/yinzhengjie/bigfile.log
    umount /tmp/yinzhengjie
    rm -rf /tmp/yinzhengjie

测试

bash 复制代码 
# 2.2 创建资源
[root@master231 configmaps]# kubectl apply -f  01-cm-conf.yaml
configmap/cm-conf created

# 2.3 查看资源
[root@master231 configmaps]# kubectl get cm 
NAME               DATA   AGE
cm-conf            4      4s
kube-root-ca.crt   1      6d8h

[root@master231 configmaps]# kubectl get cm  cm-conf 
NAME      DATA   AGE
cm-conf   4      103s

# 2.4 删除资源
[root@master231 configmaps]# kubectl delete -f 01-cm-conf.yaml 
configmap "cm-conf" deleted

Pod基于环境变量引用cm资源

创建资源

bash 复制代码 
[root@master231 configmaps]# kubectl apply -f 01-cm-conf.yaml 
[root@master231 configmaps]# kubectl get cm cm-conf

编写资源清单

yaml 复制代码 
# ./02-pods-env-cm.yaml
apiVersion: v1
kind: Pod
metadata:
  name: pods-cm-env
spec:
  containers:
  - name: c1
    image: registry.cn-hangzhou.aliyuncs.com/yinzhengjie-k8s/apps:v2

    env:
    - name: oldboyedu-SCHOOL
      # 表示值从某个地方引用
      valueFrom:
        # 值从一个cm资源引用
        configMapKeyRef:
          # 指定的是cm的资源清单的名称
          name: cm-conf
          # 指定是cm中存在的某个key
          key: school

    - name: oldboyedu-MySQL
      valueFrom:
        configMapKeyRef:
          name: cm-conf
          key: my.cnf

测试验证

bash 复制代码 
[root@master231 configmaps]# kubectl apply -f ./02-pods-env-cm.yaml
pod/pods-cm-env created

[root@master231 configmaps]# kubectl get pods pods-cm-env 
NAME          READY   STATUS    RESTARTS   AGE
pods-cm-env   1/1     Running   0          16s

[root@master231 configmaps]# kubectl exec  pods-cm-env -- env 
...
oldboyedu-SCHOOL=oldboyedu
oldboyedu-MySQL=[mysqld]
datadir=/oldboyedu/data/mysql80
basedir=/oldboyedu/softwares/mysql80
socket=/tmp/mysql80
[client]
username=linux99
password=oldboyedu

Pod基于存储卷的方式引用cm资源

编写资源清单

yaml 复制代码 
# 03-pods-volumes-cm.yaml
apiVersion: v1
kind: Pod
metadata:
  name: pods-cm-env-volumes
spec:
  # 进行存储卷引用
  volumes:
  - name: data
    # 指定configMap为存储卷类型
    configMap:
      # 指定存储卷的名称
      name: cm-conf
      # 指引引用的具体key，若不指定，则默认引用所有的key。
      items:
        # 引用cm的key
      - key: my.cnf
        # 此处可以理解为将来挂载点的文件名称
        path: mysql.txt

      - key: start.sh
        path: oldboyedu-start.sh

  containers:
  - name: c1
    image: registry.cn-hangzhou.aliyuncs.com/yinzhengjie-k8s/apps:v2

    # 进去容器就可以在/oldboyedu下面显示你引用的文件了
    volumeMounts:
    - name: data
      mountPath: /oldboy·edu

    # 这个是环境变量：进去容器的env就会显示oldboyedu-SCHOOL=oldboyedu     olboyedu-MySQL=那个资源清单
    env:
    - name: oldboyedu-SCHOOL
      valueFrom:
        configMapKeyRef:
          name: cm-conf
          key: school

    - name: oldboyedu-MySQL
      valueFrom:
        configMapKeyRef:
          name: cm-conf
          key: my.cnf

测试验证

bash 复制代码 
[root@master231 configmaps]# kubectl apply -f 03-pods-volumes-cm.yaml 
pod/pods-cm-env-volumes created

[root@master231 configmaps]# kubectl get pods -o wide pods-cm-env-volumes 
NAME                  READY   STATUS    RESTARTS   AGE   IP            NODE        NOMINATED NODE   READINESS GATES
pods-cm-env-volumes   1/1     Running   0          6s    10.100.1.70   worker232   <none>           <none>

[root@master231 configmaps]# kubectl exec -it pods-cm-env-volumes -- sh
/ # ls -l /oldboyedu/
total 0
lrwxrwxrwx    1 root     root            16 Sep 25 11:28 mysql.txt -> ..data/mysql.txt
lrwxrwxrwx    1 root     root            25 Sep 25 11:28 oldboyedu-start.sh -> ..data/oldboyedu-start.sh

/ # cat /oldboyedu/mysql.txt 
[mysqld]
datadir=/oldboyedu/data/mysql80
basedir=/oldboyedu/softwares/mysql80
socket=/tmp/mysql80
[client]
username=linux99
password=oldboyedu

/ # cat /oldboyedu/oldboyedu-start.sh 
#!/bin/sh
mkdir /tmp/yinzhengjie
mount -t tmpfs -o size=90M tmpfs /tmp/yinzhengjie/
dd if=/dev/zero of=/tmp/yinzhengjie/bigfile.log
sleep 30
rm /tmp/yinzhengjie/bigfile.log
umount /tmp/yinzhengjie
rm -rf /tmp/yinzhengjie

案例：ConfigMap实现配置中心案例

目的：统一切换nginx配置文件的端口

编写资源清单

yaml 复制代码 
# 04-deploy-cm-xiuxian.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: deploy-xiuxian
spec:
  replicas: 3
  selector:
    matchLabels:
      app: xiuxian
  template:
    metadata:
      labels:
        app: xiuxian
        version: v1

    spec:
      volumes:
      - name: data
        configMap:
          name: cm-nginx      # 引用的配置就是下面的cm-nginx
          items:
          - key: default.conf
            path: xiuxian.conf

      containers:
      - name: c1
        image: registry.cn-hangzhou.aliyuncs.com/yinzhengjie-k8s/apps:v1

        volumeMounts:
        - name: data
          mountPath: /etc/nginx/conf.d/             # 所以到时就在/etc/nginx/config.d/xiuxian.conf
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: cm-nginx
data:
  default.conf: |
    server {        # 监听端口91
        listen       91;
        listen  [::]:91;
        server_name  localhost;
        location / {
            root   /usr/share/nginx/html;
            index  index.html index.htm;
        }
        error_page   500 502 503 504  /50x.html;
        location = /50x.html {
            root   /usr/share/nginx/html;
        }
    }

测试验证

bash 复制代码 
[root@master231 configmaps]# kubectl apply -f 04-deploy-cm-xiuxian.yaml 
deployment.apps/deploy-xiuxian created
configmap/cm-nginx created
[root@master231 configmaps]# kubectl get pods -o wide
NAME                              READY   STATUS    RESTARTS   AGE   IP             NODE        NOMINATED NODE   READINESS GATES
deploy-xiuxian-84885b878c-7wv8k   1/1     Running   0          3s    10.100.2.188   worker233   <none>           <none>
deploy-xiuxian-84885b878c-lsz8z   1/1     Running   0          3s    10.100.1.90    worker232   <none>           <none>
deploy-xiuxian-84885b878c-tkcvv   1/1     Running   0          3s    10.100.2.189   worker233   <none>           <none>
[root@master231 configmaps]# curl 10.100.2.188
curl: (7) Failed to connect to 10.100.2.188 port 80 after 0 ms: Connection refused

[root@master231 configmaps]# curl 10.100.2.188:90
<!DOCTYPE html>
<html>
  <head>
    <meta charset="utf-8"/>
    <title>yinzhengjie apps v1</title>
    <style>
       div img {
          width: 900px;
          height: 600px;
          margin: 0;
       }
    </style>
  </head>
  <body>
    <h1 style="color: green">凡人修仙传 v1 </h1>
    <div>
      <img src="1.jpg">
    <div>
  </body>
</html>

# 注意：如需再次切换，需要删除并重启pod资源清单
相关推荐
DeeplyMind3 小时前
第23章 ROCm虚拟化栈Docker化实战
运维·docker·容器
吹牛不交税4 小时前
.netcore项目部署在ubuntu22.04虚拟机的docker中的过程记录
docker·容器·.netcore
人间打气筒（Ada）5 小时前
Kubernetes核心技术-namespace & label
云原生·kubernetes·云计算·k8s·devops·namespace·label
谢斯6 小时前
[AFFine]文档系统,docker搭建
运维·docker·容器
礼拜天没时间.12 小时前
Docker Registry私有仓库搭建与使用
java·运维·docker·云原生·容器·centos
nix.gnehc12 小时前
零基础部署K8s单节点集群：一键脚本实现快速落地
云原生·容器·kubernetes
zhu620197612 小时前
Postgres数据库docker快速安装
数据库·docker·容器
数据知道13 小时前
PostgreSQL：如何在容器中部署 Crunchy Data Operator（Kubernetes 运维）
运维·postgresql·kubernetes
only_Klein14 小时前
Jenkinsfile流水线设计解析
kubernetes·jenkins·argocd·cicd
热门推荐
01GitHub 镜像站点02从零搭建一个 PHP 登录注册系统（含完整源码）03使用 1panel面板 部署 php网站04AI Agent 平台横评：ZeroClaw vs OpenClaw vs Nanobot05openClaw安装飞书插件｜核心踩坑：spawn EINVAL 错误终极解决指南06MIUI显示/隐藏5G开关的方法，信号弱时开启手机Wifi通话方法07Claude Code + GLM4.7 避坑指南：解决 Unable to connect to Anthropic services08让你的代码更整洁：10 个必知的 Kotlin 扩展函数09如何解决 OpenClaw “Pairing required” 报错：两种官方解决方案详解10全面体验 Grok API 中转站（2025 · Grok 4 系列最新版）