Kubernetes控制器详解

华子目录

• 什么是控制器
• 控制器常用类型
• `replicaset`控制器
• • `replicaset`功能
  • • `replicaset`参数说明
    • `replicaset`示例
• `deployment`控制器
• • `deployment`控制器的功能
  • • 典型的应用场景
    • `Deployment`控制器`示例`
    • • 1.版本迭代
      • 2.版本回滚
      • 3.滚动更新策略
      • 4.暂停及恢复
• `daemonset`控制器
• • `daemonset`功能
  • `daemonset`的典型用法
  • `daemonset`示例
• `job`控制器
• • `job`控制器功能
  • `job`的特点
  • `job`控制器示例
• `cronjob`控制器
• • `cronjob`控制器功能
  • `cronjob`控制器示例

什么是控制器

官方文档：https://v1-30.docs.kubernetes.io/zh-cn/docs/concepts/workloads/controllers/

控制器也是管理pod的一种手段

• 自主式pod：pod退出或意外关闭后不会被重新创建
• 控制器管理的 Pod：在控制器的生命周期里，始终要维持 Pod 的副本数目

Pod控制器是管理pod的中间层，使用Pod控制器之后，只需要告诉Pod控制器，想要多少个什么样的
Pod就可以了，它会创建出满足条件的Pod并确保每一个Pod资源处于用户期望的目标状态。如果Pod资源在运行中出现故障，它会基于指定策略重新编排Pod

当建立控制器后，会把期望值写入etcd，k8s中的apiserver检索etcd中我们保存的期望状态，并对比pod的当前状态，如果出现差异，代码自驱动立即恢复

控制器常用类型

控制器名称	控制器用途
Replication Controller	比较原始的pod控制器，已经被废弃，由ReplicaSet代替
ReplicaSet	ReplicaSet 确保任何时间都有指定数量的 Pod 副本在运行
Deployment	一个 Deployment 为 Pod 和 ReplicaSet 提供声明式的更新能力
DaemonSet	DaemonSet 确保全指定节点上运行一个 Pod 的副本
StatefulSet	StatefulSet 是用来管理有状态应用的工作负载 API 对象
Job	执行批处理任务，仅执行一次任务，保证任务的一个或多个Pod成功结束
CronJob	Cron Job 创建基于时间调度的 Jobs
HPA全称Horizontal Pod Autoscaler	根据资源利用率自动调整service中Pod数量，实现Pod水平自动缩放

replicaset控制器

replicaset功能

• ReplicaSet 是下一代的 Replication Controller，官方推荐使用ReplicaSet
• ReplicaSet和Replication Controller的唯一区别是选择器的支持，ReplicaSet支持新的基于集合的选择器需求
• ReplicaSet确保任何时间都有指定数量的 Pod 副本在运行
• 虽然 ReplicaSets 可以独立使用，但今天它主要被Deployments 用作协调 Pod 创建、删除和更新的机制

replicaset参数说明

参数名称	字段类型	参数说明
spec	object	详细定义对象，固定值就写spec
spec.replicas	int	指定维护pod数量
spec.selector	object	Selector是对pod的标签查询，与pod数量匹配
spec.selector.matchLabels	string	指定Selector查询标签的名称和值，以key：value方式指定
spec.template	object	指定对pod的描述信息，比如lab标签，运行容器的信息等
spec.template.metadata	object	指定pod属性
spec.template.metadata.labels	string	指定pod标签
spec.template.spec	object	详细定义对象
spec.template.spec.containers	list	Spec对象的容器列表定义
spec.template.spec.containers.name	string	指定容器名称
spec.template.spec.containers.image	string	指定容器镜像

replicaset示例

[root@k8s-master ~]# kubectl create deployment replicaset --image myapp:v1 --dry-run=client -o yaml > replicaset.yml


[root@k8s-master ~]# vim replicaset.yml
[root@k8s-master ~]# cat replicaset.yml
apiVersion: apps/v1
kind: ReplicaSet   #ReplicaSet控制器
metadata:
  labels:
    app: myapp     #标签为app=myapp
  name: replicaset   #指定ReplicaSet控制器名称，一定小写，如果出现大写报错
spec:
  replicas: 2   #指定维护pod数量为2
  selector:  #指定检测匹配方式
    matchLabels:   #指定匹配方式为标签匹配
      app: myapp   #指定匹配的标签为app=myapp
  template:   #模板：当pod数量不足时，会根据下面的模板创建pod副本
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - image: myapp:v1
        name: myapp


[root@k8s-master ~]# kubectl apply -f replicaset.yml
replicaset.apps/replicaset created


[root@k8s-master ~]# kubectl get pods -o wide
NAME               READY   STATUS    RESTARTS   AGE   IP            NODE            NOMINATED NODE   READINESS GATES
replicaset-82kts   1/1     Running   0          12s   10.244.2.32   k8s-node2.org   <none>           <none>
replicaset-vn55z   1/1     Running   0          12s   10.244.1.23   k8s-node1.org   <none>           <none>


[root@k8s-master ~]# kubectl get pods --show-labels
NAME               READY   STATUS    RESTARTS   AGE   LABELS
replicaset-82kts   1/1     Running   0          60s   app=myapp
replicaset-vn55z   1/1     Running   0          60s   app=myapp

• replicaset是通过标签匹配pod的

[root@k8s-master ~]# kubectl label pods replicaset-82kts app=huazi --overwrite
pod/replicaset-82kts labeled


[root@k8s-master ~]# kubectl get pods --show-labels
NAME               READY   STATUS    RESTARTS   AGE     LABELS
replicaset-82kts   1/1     Running   0          3m43s   app=huazi
replicaset-clsjq   1/1     Running   0          9s      app=myapp   #重新开启的pod
replicaset-vn55z   1/1     Running   0          3m43s   app=myapp

• 恢复标签后，发现新创建的pod没了

[root@k8s-master ~]# kubectl label pods replicaset-82kts app=myapp --overwrite
pod/replicaset-82kts labeled
[root@k8s-master ~]# kubectl get pods --show-labels
NAME               READY   STATUS    RESTARTS   AGE    LABELS
replicaset-82kts   1/1     Running   0          7m2s   app=myapp
replicaset-vn55z   1/1     Running   0          7m2s   app=myapp

• replicaset自动控制pod数量

deployment控制器

deployment控制器的功能

• 为了更好的解决服务编排的问题，kuberenetes在v1.2版本开始，引入了Deployment控制器。
• Deployment控制器并不直接管理pod，而是通过管理ReplicaSet来间接管理pod
• Deployment管理ReplicaSet，ReplicaSet管理pod
• Deployment为Pod和ReplicaSet提供了一个申明式的定义方法
• 在Deployment中ReplicaSet相当于一个版本

典型的应用场景

• 用来创建Pod和ReplicaSet
• 滚动更新和回滚
• 扩容和缩容
• 暂停与恢复

Deployment控制器示例

[root@k8s-master ~]# mkdir deployment
[root@k8s-master ~]# cd deployment/
[root@k8s-master deployment]# kubectl create deployment deployment --image myapp:v1 --dry-run=client -o yaml > deployment.yml
[root@k8s-master deployment]# ls
deployment.yml
[root@k8s-master deployment]# vim deployment.yml
[root@k8s-master deployment]# cat deployment.yml
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: myapp   #便签app=myapp
  name: deployment  #指定Deployment控制器名称，一定小写，如果出现大写报错
spec:
  replicas: 4   #控制器中pod数量为4个
  selector:
    matchLabels:
      app: myapp
  template:  #模板：当pod数量不足时，会根据下面的模板创建pod副本
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - image: myapp:v1
        name: myapp


[root@k8s-master deployment]# kubectl apply -f deployment.yml
deployment.apps/deployment created
[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE   IP           NODE            NOMINATED NODE   READINESS GATES
deployment-5d886954d4-2j8jc   1/1     Running   0          13s   10.244.1.5   k8s-node1.org   <none>           <none>
deployment-5d886954d4-8dpr7   1/1     Running   0          13s   10.244.1.4   k8s-node1.org   <none>           <none>
deployment-5d886954d4-9zn5b   1/1     Running   0          13s   10.244.2.7   k8s-node2.org   <none>           <none>
deployment-5d886954d4-zcnxc   1/1     Running   0          13s   10.244.2.6   k8s-node2.org   <none>           <none>


[root@k8s-master deployment]# kubectl get pods --show-labels
NAME                          READY   STATUS    RESTARTS   AGE   LABELS
deployment-5d886954d4-2j8jc   1/1     Running   0          65s   app=myapp,pod-template-hash=5d886954d4
deployment-5d886954d4-8dpr7   1/1     Running   0          65s   app=myapp,pod-template-hash=5d886954d4
deployment-5d886954d4-9zn5b   1/1     Running   0          65s   app=myapp,pod-template-hash=5d886954d4
deployment-5d886954d4-zcnxc   1/1     Running   0          65s   app=myapp,pod-template-hash=5d886954d4

1.版本迭代

[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE    IP           NODE            NOMINATED NODE   READINESS GATES
deployment-5d886954d4-2j8jc   1/1     Running   0          2m1s   10.244.1.5   k8s-node1.org   <none>           <none>
deployment-5d886954d4-8dpr7   1/1     Running   0          2m1s   10.244.1.4   k8s-node1.org   <none>           <none>
deployment-5d886954d4-9zn5b   1/1     Running   0          2m1s   10.244.2.7   k8s-node2.org   <none>           <none>
deployment-5d886954d4-zcnxc   1/1     Running   0          2m1s   10.244.2.6   k8s-node2.org   <none>           <none>

#访问第一个pod中的ip，我们发现是v1版本
[root@k8s-master deployment]# curl 10.244.1.5
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>


[root@k8s-master deployment]# kubectl describe deployments.apps deployment
Name:                   deployment
Namespace:              default  #命名空间
CreationTimestamp:      Tue, 15 Oct 2024 10:53:21 -0400
Labels:                 app=myapp
Annotations:            deployment.kubernetes.io/revision: 1
Selector:               app=myapp
Replicas:               4 desired | 4 updated | 4 total | 4 available | 0 unavailable
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge  ##默认每次更新25%
Pod Template:
  Labels:  app=myapp
  Containers:
   myapp:
    Image:         myapp:v1
    Port:          <none>
    Host Port:     <none>
    Environment:   <none>
    Mounts:        <none>
  Volumes:         <none>
  Node-Selectors:  <none>
  Tolerations:     <none>
Conditions:
  Type           Status  Reason
  ----           ------  ------
  Available      True    MinimumReplicasAvailable
  Progressing    True    NewReplicaSetAvailable
OldReplicaSets:  <none>
NewReplicaSet:   deployment-5d886954d4 (4/4 replicas created)
Events:
  Type    Reason             Age    From                   Message
  ----    ------             ----   ----                   -------
  Normal  ScalingReplicaSet  3m21s  deployment-controller  Scaled up replica set deployment-5d886954d4 to 4



[root@k8s-master deployment]# vim deployment.yml

#我们发现，每次apply后，pod的ip都会变
[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE   IP           NODE            NOMINATED NODE   READINESS GATES
deployment-7f4786db9c-5mxc2   1/1     Running   0          26s   10.244.1.7   k8s-node1.org   <none>           <none>
deployment-7f4786db9c-hbfmh   1/1     Running   0          27s   10.244.1.6   k8s-node1.org   <none>           <none>
deployment-7f4786db9c-pswjb   1/1     Running   0          27s   10.244.2.8   k8s-node2.org   <none>           <none>
deployment-7f4786db9c-xpqlf   1/1     Running   0          26s   10.244.2.9   k8s-node2.org   <none>           <none>

#测试：发现版本成为了V2
[root@k8s-master deployment]# curl 10.244.1.7
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

• 注意：更新的过程是重新建立一个版本的RS，新版本的RS会把pod重建，然后把老版本的RS回收

2.版本回滚

[root@k8s-master deployment]# vim deployment.yml

[root@k8s-master deployment]# kubectl apply -f deployment.yml
deployment.apps/deployment configured

#我们发现，每次apply后，pod的ip都会变
[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE   IP            NODE            NOMINATED NODE   READINESS GATES
deployment-5d886954d4-f4wsl   1/1     Running   0          6s    10.244.2.10   k8s-node2.org   <none>           <none>
deployment-5d886954d4-j8qkt   1/1     Running   0          6s    10.244.1.8    k8s-node1.org   <none>           <none>
deployment-5d886954d4-n5v9v   1/1     Running   0          5s    10.244.1.9    k8s-node1.org   <none>           <none>
deployment-5d886954d4-vj64d   1/1     Running   0          5s    10.244.2.11   k8s-node2.org   <none>           <none>


#测试
[root@k8s-master deployment]# curl 10.244.2.10
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

3.滚动更新策略

[root@k8s-master deployment]# vim deployment.yml
[root@k8s-master deployment]# cat deployment.yml
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: myapp
  name: deployment
spec:
  minReadySeconds: 5   #最小就绪时间，指定pod每隔多久更新一次
  replicas: 4
  strategy:   #指定更新策略
    rollingUpdate:
      maxSurge: 1    #比定义pod数量多几个
      maxUnavailable: 0   #比定义pod个数少几个
  selector:
    matchLabels:
      app: myapp
  template:
    metadata:
      labels:
        app: myapp
    spec:
      containers:
      - image: myapp:v1
        name: myapp


[root@k8s-master deployment]# kubectl apply -f deployment.yml
deployment.apps/deployment created


[root@k8s-master deployment]# kubectl get all
NAME                              READY   STATUS    RESTARTS   AGE
pod/deployment-5d886954d4-67kjg   1/1     Running   0          43s
pod/deployment-5d886954d4-crrwc   1/1     Running   0          43s
pod/deployment-5d886954d4-ksfnh   1/1     Running   0          43s
pod/deployment-5d886954d4-nql72   1/1     Running   0          43s

NAME                 TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
service/huazi        ClusterIP   10.97.207.164   <none>        8080/TCP   6d22h
service/kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP    11d

NAME                         READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/deployment   4/4     4            4           43s

NAME                                    DESIRED   CURRENT   READY   AGE
replicaset.apps/deployment-5d886954d4   4         4         4       43s

4.暂停及恢复

• 在实际生产环境中我们做的变更可能不止一处，当修改了一处后，如果执行变更就直接触发了
• 我们期望是当我们把所有修改都搞定后，一次触发

#先暂停
[root@k8s-master deployment]# kubectl rollout pause deployment deployment
deployment.apps/deployment paused


[root@k8s-master deployment]# vim deployment.yml

[root@k8s-master deployment]# kubectl apply -f deployment.yml
deployment.apps/deployment configured

[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE     IP            NODE            NOMINATED NODE   READINESS GATES
deployment-5d886954d4-4wz8z   1/1     Running   0          9m10s   10.244.2.14   k8s-node2.org   <none>           <none>
deployment-5d886954d4-67kjg   1/1     Running   0          18m     10.244.2.13   k8s-node2.org   <none>           <none>
deployment-5d886954d4-crrwc   1/1     Running   0          18m     10.244.1.10   k8s-node1.org   <none>           <none>
deployment-5d886954d4-fz7t4   1/1     Running   0          9m10s   10.244.1.12   k8s-node1.org   <none>           <none>
deployment-5d886954d4-ksfnh   1/1     Running   0          18m     10.244.1.11   k8s-node1.org   <none>           <none>
deployment-5d886954d4-nql72   1/1     Running   0          18m     10.244.2.12   k8s-node2.org   <none>           <none>

#发现版本并没有被更新
[root@k8s-master deployment]# curl 10.244.2.14
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

#也没有发现资源限制
[root@k8s-master deployment]# kubectl describe pods deployment-5d886954d4-4wz8z
Name:             deployment-5d886954d4-4wz8z
Namespace:        default
Priority:         0
Service Account:  default
Node:             k8s-node2.org/172.25.254.20
Start Time:       Tue, 15 Oct 2024 11:35:29 -0400
Labels:           app=myapp
                  pod-template-hash=5d886954d4
Annotations:      <none>
Status:           Running
IP:               10.244.2.14
IPs:
  IP:           10.244.2.14
Controlled By:  ReplicaSet/deployment-5d886954d4
Containers:
  myapp:
    Container ID:   docker://9430d7c64e16a356882389763c74913dfe9081fb6c8534fdc760f39bb7d8fe8b
    Image:          myapp:v1
    Image ID:       docker-pullable://myapp@sha256:9eeca44ba2d410e54fccc54cbe9c021802aa8b9836a0bcf3d3229354e4c8870e
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Tue, 15 Oct 2024 11:35:29 -0400
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-6qzsv (ro)
Conditions:
  Type                        Status
  PodReadyToStartContainers   True
  Initialized                 True
  Ready                       True
  ContainersReady             True
  PodScheduled                True
Volumes:
  kube-api-access-6qzsv:
    Type:                    Projected (a volume that contains injected data from multiple sources)
    TokenExpirationSeconds:  3607
    ConfigMapName:           kube-root-ca.crt
    ConfigMapOptional:       <nil>
    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  12m   default-scheduler  Successfully assigned default/deployment-5d886954d4-4wz8z to k8s-node2.org
  Normal  Pulled     12m   kubelet            Container image "myapp:v1" already present on machine
  Normal  Created    12m   kubelet            Created container myapp
  Normal  Started    12m   kubelet            Started container myapp

[root@k8s-master deployment]# kubectl rollout history deployment deployment
deployment.apps/deployment
REVISION  CHANGE-CAUSE
1         <none>

#解除暂停
[root@k8s-master deployment]# kubectl rollout resume deployment deployment
deployment.apps/deployment resumed


[root@k8s-master deployment]# kubectl rollout history deployment deployment
deployment.apps/deployment
REVISION  CHANGE-CAUSE
1         <none>
2         <none>

#发现pod中的ip也变了
[root@k8s-master deployment]# kubectl get pods -o wide
NAME                          READY   STATUS    RESTARTS   AGE   IP            NODE            NOMINATED NODE   READINESS GATES
deployment-79fcdf95d5-2xh8f   1/1     Running   0          43s   10.244.1.15   k8s-node1.org   <none>           <none>
deployment-79fcdf95d5-7s79k   1/1     Running   0          62s   10.244.2.15   k8s-node2.org   <none>           <none>
deployment-79fcdf95d5-b8459   1/1     Running   0          55s   10.244.1.14   k8s-node1.org   <none>           <none>
deployment-79fcdf95d5-c7dsd   1/1     Running   0          37s   10.244.2.17   k8s-node2.org   <none>           <none>
deployment-79fcdf95d5-cncld   1/1     Running   0          49s   10.244.2.16   k8s-node2.org   <none>           <none>
deployment-79fcdf95d5-jcndp   1/1     Running   0          67s   10.244.1.13   k8s-node1.org   <none>           <none>


#发现版本更新为V2
[root@k8s-master deployment]# curl 10.244.1.15
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>

#也发现了资源限制
[root@k8s-master deployment]# kubectl describe pods deployment-79fcdf95d5-2xh8f
Name:             deployment-79fcdf95d5-2xh8f
Namespace:        default
Priority:         0
Service Account:  default
Node:             k8s-node1.org/172.25.254.10
Start Time:       Tue, 15 Oct 2024 11:50:08 -0400
Labels:           app=myapp
                  pod-template-hash=79fcdf95d5
Annotations:      <none>
Status:           Running
IP:               10.244.1.15
IPs:
  IP:           10.244.1.15
Controlled By:  ReplicaSet/deployment-79fcdf95d5
Containers:
  myapp:
    Container ID:   docker://54b5f21974e014328d77236a2b57b84afd66afff0a158768eb6928f513f9ba04
    Image:          myapp:v2
    Image ID:       docker-pullable://myapp@sha256:5f4afc8302ade316fc47c99ee1d41f8ba94dbe7e3e7747dd87215a15429b9102
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Tue, 15 Oct 2024 11:50:08 -0400
    Ready:          True
    Restart Count:  0
    Limits:
      cpu:     500m
      memory:  200Mi
    Requests:
      cpu:        500m
      memory:     200Mi
    Environment:  <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-gb8n4 (ro)
Conditions:
  Type                        Status
  PodReadyToStartContainers   True
  Initialized                 True
  Ready                       True
  ContainersReady             True
  PodScheduled                True
Volumes:
  kube-api-access-gb8n4:
    Type:                    Projected (a volume that contains injected data from multiple sources)
    TokenExpirationSeconds:  3607
    ConfigMapName:           kube-root-ca.crt
    ConfigMapOptional:       <nil>
    DownwardAPI:             true
QoS Class:                   Guaranteed
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  2m24s  default-scheduler  Successfully assigned default/deployment-79fcdf95d5-2xh8f to k8s-node1.org
  Normal  Pulled     2m24s  kubelet            Container image "myapp:v2" already present on machine
  Normal  Created    2m24s  kubelet            Created container myapp
  Normal  Started    2m24s  kubelet            Started container myapp

• 回收

[root@k8s-master deployment]# kubectl delete -f deployment.yml
deployment.apps "deployment" deleted

daemonset控制器

daemonset功能

DaemonSet确保全部（或者某些）节点上运行一个Pod。当有节点加入集群时， 也会为他们新增一个Pod ，当有节点从集群移除时，这些 Pod也被回收。删除 DaemonSet 将会删除它创建的所有Pod

daemonset的典型用法

• 在每个节点上运行集群存储DaemonSet，例如glusterd、ceph。
• 在每个节点上运行日志收集DaemonSet，例如fluentd、logstash。
• 在每个节点上运行监控DaemonSet，例如Prometheus Node Exporter、zabbix agent等
• 一个简单的用法是在所有的节点上都启动一个DaemonSet，将被作为每种类型的daemon使用
• 一个稍微复杂的用法是单独对每种daemon类型使用多个DaemonSet，但具有不同的标志， 并且对不同硬件类型具有不同的内存、CPU要求

daemonset示例

• 由于daemonset不能create，所以这里我们进行手写

[root@k8s-master ~]# mkdir daemonset
[root@k8s-master ~]# cd daemonset/

[root@k8s-master daemonset]# vim daemonset.yml
[root@k8s-master daemonset]# cat daemonset.yml
apiVersion: apps/v1
kind: DaemonSet
metadata:
  labels:
    app: myapp
  name: daemonset
spec:
  selector:
    matchLabels:
      app: myapp
  template:
    metadata:
      labels:
        app: myapp
    spec:
      tolerations:   #设置对污点的容忍
      - effect: NoSchedule
        operator: Exists
      containers:
        - image: myapp:v1
          name: myapp


[root@k8s-master daemonset]# kubectl apply -f daemonset.yml
daemonset.apps/daemonset created


#发现master上也可以运行pod
[root@k8s-master daemonset]# kubectl get pods -o wide
NAME              READY   STATUS              RESTARTS   AGE   IP            NODE             NOMINATED NODE   READINESS GATES
daemonset-9qznl   1/1     Running             0          13s   10.244.2.18   k8s-node2.org    <none>           <none>
daemonset-h7ggv   0/1     ContainerCreating   0          13s   <none>        k8s-master.org   <none>           <none>
daemonset-m78l6   1/1     Running             0          13s   10.244.1.16   k8s-node1.org    <none>           <none>

• 回收

[root@k8s-master daemonset]# kubectl delete -f daemonset.yml
daemonset.apps "daemonset" deleted

job控制器

job控制器功能

job，主要用于负责批量处理（一次要处理指定数量任务）短暂的一次性（每个任务仅运行一次就结束）任务

job的特点

• 当job创建的pod执行成功结束时，job将记录成功结束的pod数量
• 当成功结束的pod达到指定的数量时，job将完成执行

job控制器示例

[root@k8s-master ~]# mkdir job
[root@k8s-master ~]# cd job/

cronjob控制器

cronjob控制器功能

• CronJob 创建基于时间调度的Jobs
• CronJob控制器以Job控制器资源为其管控对象，并借助它管理pod资源对象
• CronJob可以以类似于Linux操作系统的周期性任务作业计划的方式控制其运行时间点及重复运行的方式。
• CronJob可以在特定的时间点(反复的)去运行job任务

cronjob控制器示例