21-kubeadm底层实现之静态Pod

一、静态Pod概述

所谓的静态pod就是kubelet自己监视的一个目录，如果这个目录有Pod资源清单，就直接会在当前节点上创建该Pod。也就是说不基于APIServer就可以直接创建Pod。

静态Pod仅对Pod类型的资源有效，其他资源无视。

静态Pod创建的资源，后缀都会加一个当前节点的名称

二、静态Pod路径概述

vim  /var/lib/kubelet/config.yaml 
...
staticPodPath: /etc/kubernetes/manifests

温馨提示:
	(1)静态Pod是由kubelet启动时通过"staticPodPath"配置参数指定路径
	(2)静态Pod创建的Pod名称会自动加上kubelet节点的主机名，比如"-k8s231.oldboyedu.com"，会忽略"nodeName"字段哟;
	(3)静态Pod的创建并不依赖API-Server，而是直接基于kubelet所在节点来启动Pod;
	(4)静态Pod的删除只需要将其从staticPodPath指定的路径移除即可;
	(5)静态Pod路径仅对Pod资源类型有效，其他类型资源将不被创建哟
	(6)咱们的kubeadm部署方式就是基于静态Pod部署的哟;

三、参考案例

#查看静态资源文件的路径
[root@master231 ~]# grep staticPodPath /var/lib/kubelet/config.yaml 
staticPodPath: /etc/kubernetes/manifests

#进入目录
[root@master231 ~]# cd /etc/kubernetes/manifests/
[root@master231 manifests]# 
[root@master231 manifests]# ll
total 24
drwxr-xr-x 2 root root 4096 Apr  7 11:00 ./
drwxr-xr-x 4 root root 4096 Apr  7 11:00 ../
-rw------- 1 root root 2280 Apr  7 11:00 etcd.yaml
-rw------- 1 root root 4025 Apr  7 11:00 kube-apiserver.yaml
-rw------- 1 root root 3546 Apr  7 11:00 kube-controller-manager.yaml
-rw------- 1 root root 1465 Apr  7 11:00 kube-scheduler.yaml

四、查看

我们不难发现，此目录下为k8s架构的各个组件，我们可以查看一番,其中tier 是一个标签（label），用于对资源进行分类和分层。

[root@master231 manifests]# head * 
==> etcd.yaml <==
apiVersion: v1
kind: Pod
metadata:
  annotations:
    kubeadm.kubernetes.io/etcd.advertise-client-urls: https://10.0.0.231:2379
  creationTimestamp: null
  labels:
    component: etcd
    tier: control-plane
  name: etcd

==> kube-apiserver.yaml <==
apiVersion: v1
kind: Pod
metadata:
  annotations:
    kubeadm.kubernetes.io/kube-apiserver.advertise-address.endpoint: 10.0.0.231:6443
  creationTimestamp: null
  labels:
    component: kube-apiserver
    tier: control-plane
  name: kube-apiserver

==> kube-controller-manager.yaml <==
apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: null
  labels:
    component: kube-controller-manager
    tier: control-plane
  name: kube-controller-manager
  namespace: kube-system
spec:

==> kube-scheduler.yaml <==
apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: null
  labels:
    component: kube-scheduler
    tier: control-plane
  name: kube-scheduler
  namespace: kube-system
spec:

我们指定名称空间和标签来查看下都运行了哪些Pod

[root@master231 manifests]# kubectl get pods -o wide -n kube-system -l  tier=control-plane
NAME                                READY   STATUS    RESTARTS       AGE    IP           NODE        NOMINATED NODE   READINESS GATES
etcd-master231                      1/1     Running   1 (2d4h ago)   2d5h   10.0.0.231   master231   <none>           <none>
kube-apiserver-master231            1/1     Running   1 (2d4h ago)   2d5h   10.0.0.231   master231   <none>           <none>
kube-controller-manager-master231   1/1     Running   1 (2d4h ago)   2d5h   10.0.0.231   master231   <none>           <none>
kube-scheduler-master231            1/1     Running   1 (2d4h ago)   2d5h   10.0.0.231   master231   <none>           <none>

五、基于静态Pod修改svc的NodePort端口范围

1.默认的端口范围是30000~32767

[root@master231 service]# cat 02-svc-NodePort-xiuxian.yaml 
apiVersion: v1
kind: Service
metadata:
  labels:
    apps: xiuxian
  name: svc-xiuxian-nodeport
spec:
  type: NodePort
  ports:
  - port: 90
    protocol: TCP
    targetPort: 80
    # 声明worker节点的转发的端口，默认的有效范围是:  30000-32767
    # nodePort: 30080
    nodePort: 8080
  selector:
    version: v1
    
   #报错提示我们了
[root@master231 service]# kubectl apply -f  02-svc-NodePort-xiuxian.yaml 
The Service "svc-xiuxian-nodeport" is invalid: spec.ports[0].nodePort: Invalid value: 8080: provided port is not in the valid range. The range of valid ports is 30000-32767

2.修改NodePort的默认端口范围

[root@master231 ~]# vim /etc/kubernetes/manifests/kube-apiserver.yaml 
 ...
 spec:
  containers:
  - command:
    - kube-apiserver
    - --service-node-port-range=3000-50000  # 进行添加这一行即可
    ...

3.移动镜像Pod目录的资源清单使Pod重启，配置生效

[root@master231 manifests]# pwd
/etc/kubernetes/manifests
[root@master231 manifests]# 
[root@master231 manifests]# mv kube-apiserver.yaml /opt/
[root@master231 manifests]# 
[root@master231 manifests]# mv /opt/kube-apiserver.yaml ./

4.再次查看api-server的Pod启动时间

#发现kube-apiserver-master231已重新运行
[root@master231 manifests]# kubectl get pods -o wide -n kube-system -l  tier=control-plane
NAME                                READY   STATUS    RESTARTS       AGE    IP           NODE        NOMINATED NODE   READINESS GATES
etcd-master231                      1/1     Running   1 (2d5h ago)   2d5h   10.0.0.231   master231   <none>           <none>
kube-apiserver-master231            1/1     Running   0              57s    10.0.0.231   master231   <none>           <none>
kube-controller-manager-master231   1/1     Running   2 (85s ago)    2d5h   10.0.0.231   master231   <none>           <none>
kube-scheduler-master231            1/1     Running   2 (85s ago)    2d5h   10.0.0.231   master231   <none>           <none>

5.再次测试验证svc的NodePort端口范围

[root@master231 service]# cat 02-svc-NodePort-xiuxian.yaml 
apiVersion: v1
kind: Service
metadata:
  labels:
    apps: xiuxian
  name: svc-xiuxian-nodeport
spec:
  type: NodePort
  ports:
  - port: 90
    protocol: TCP
    targetPort: 80
    # 声明worker节点的转发的端口，默认的有效范围是:  30000-32767
    # nodePort: 30080
    nodePort: 8080
  selector:
    version: v1
[root@master231 service]# 
[root@master231 service]# kubectl apply -f  02-svc-NodePort-xiuxian.yaml 
service/svc-xiuxian-nodeport configured
[root@master231 service]# 
[root@master231 service]# kubectl get -f 02-svc-NodePort-xiuxian.yaml
NAME                   TYPE       CLUSTER-IP     EXTERNAL-IP   PORT(S)       AGE
svc-xiuxian-nodeport   NodePort   10.200.21.38   <none>        90:8080/TCP   23h

posted @ 2025-04-09 21:36 丁志岩阅读(25) 评论(0) 收藏举报

刷新页面返回顶部

丁志岩的运维之路