008.kubernets的调度系统之标签选择器

一 Kubernetes 调度简介

除了让 kubernetes 集群调度器自动为 pod 资源选择某个节点(默认调度考虑的是资源足够,并且 load 尽量平均),有些情况我们希望能更多地控制 pod 应该如何调度。比如,集群中有些机器的配置更好( SSD,更好的内存等),我们希望比较核心的服务(比如说数据库)运行在上面;或者某两个服务的网络传输很频繁,我们希望它们最好在同一台机器上,或者同一个机房。有一些特殊的应用,我们就希望它们能运行在我们指定的节点上,还有些情况下,我们希望某个应用在所有的节点上都能运行一份。

针对不同的应用场景,kubernetes内置了多种调度方式可供选择。包括标签选择器,daemonsets,节点亲和性,pod亲和性,污点与容忍等。

二 标签选择器

这种方式其实就是我们最常用的使用label的方式,给某一个node打上特定的标签,然后在启动pod的时候,通过nodeSelector指定要调度到的node节点的标签。

2.1  nodeSelector打标签

[ ingress]# kubectl get nodes

NAME              STATUS   ROLES    AGE     VERSION
192.168.132.131   Ready    master   3d18h   v1.17.0
192.168.132.132   Ready    <none>   3d18h   v1.17.0
192.168.132.133   Ready    <none>   3d18h   v1.17.0

ingress]# kubectl get pods  -n ingress-nginx -o wide

NAME                                        READY   STATUS    RESTARTS   AGE   IP                NODE              NOMINATED NODE   READINESS GATES
nginx-ingress-controller-5c6985f9cc-wkngv   1/1     Running   0          22h   192.168.132.132   192.168.132.132   <none>           <none>

[ ingress]# kubectl get nodes --show-labels

NAME              STATUS   ROLES    AGE     VERSION   LABELS
192.168.132.131   Ready    master   3d18h   v1.17.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=192.168.132.131,kubernetes.io/os=linux,node-role.kubernetes.io/master=
192.168.132.132   Ready    <none>   3d18h   v1.17.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=192.168.132.132,kubernetes.io/os=linux
192.168.132.133   Ready    <none>   3d18h   v1.17.0   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=192.168.132.133,kubernetes.io/os=linux

默认为节点打的标签:系统版本,节点名称,系统架构

[ ingress]# vi nginx-controller.yaml 

nodeSelector:
        kubernetes.io/os: linux

这个标签对于目前的条件,没有任何意义

2.2 重新添加一个唯一标签

nodeSelector:
        kubernetes.io/os: linux
        kubernetes.io/hostname: 192.168.132.133    #这是一个危矣标签

标签是可以是多个,而且是逻辑与的关系,即当所有的标签都满足,才能成立,如果都不满足,就会pengding

[ ingress]# kubectl apply -f nginx-controller.yaml

namespace/ingress-nginx unchanged
configmap/nginx-configuration unchanged
configmap/tcp-services unchanged
configmap/udp-services unchanged
serviceaccount/nginx-ingress-serviceaccount unchanged
clusterrole.rbac.authorization.k8s.io/nginx-ingress-clusterrole unchanged
role.rbac.authorization.k8s.io/nginx-ingress-role unchanged
rolebinding.rbac.authorization.k8s.io/nginx-ingress-role-nisa-binding unchanged
clusterrolebinding.rbac.authorization.k8s.io/nginx-ingress-clusterrole-nisa-binding unchanged
deployment.apps/nginx-ingress-controller configured
limitrange/ingress-nginx configured

[ ingress]# kubectl get pods  -n ingress-nginx -o wide

NAME                                        READY   STATUS              RESTARTS   AGE   IP                NODE              NOMINATED NODE   READINESS GATES
nginx-ingress-controller-5c6985f9cc-wkngv   1/1     Running             0          22h   192.168.132.132   192.168.132.132   <none>           <none>
nginx-ingress-controller-5cffd956bf-dm9qf   0/1     ContainerCreating   0          4s    192.168.132.133   192.168.132.133   <none>           <none>

[ ingress]# kubectl get pods  -n ingress-nginx -o wide

NAME                                        READY   STATUS    RESTARTS   AGE    IP                NODE              NOMINATED NODE   READINESS GATES
nginx-ingress-controller-5cffd956bf-dm9qf   1/1     Running   0          4m3s   192.168.132.133   192.168.132.133   <none>           <none>

然后就运行在了192.168.132.133上

2.3 无满足标签

apiVersion: apps/v1
kind: Deployment
metadata:
    name: busybox
    namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      name: busybox
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 10%
      maxUnavailable: 0
  template:
    metadata:
      labels:
        name: busybox
    spec:
      nodeSelector:
        aaa: bbb
      containers:
      - name: busybox
        image: busybox
        command:
          - /bin/sh
          - -c
          - "sleep 3600"

[ deployment]# kubectl apply -f busybox-deployment.yaml

deployment.apps/busybox configured

[ deployment]# kubectl get pods

busybox-546555c84-2psbb             1/1     Running   13         24h
busybox-674bd96f74-m4fst            0/1     Pending   0          13s

[ deployment]# kubectl describe pods busybox-674bd96f74-m4fst

Type     Reason            Age        From               Message
  ----     ------            ----       ----               -------
  Warning  FailedScheduling  <unknown>  default-scheduler  0/3 nodes are available: 3 node(s) didn‘t match node selector.
  Warning  FailedScheduling  <unknown>  default-scheduler  0/3 nodes are available: 3 node(s) didn‘t match node selector.

一直pengding状态,是因为没有匹配的标签

2.4 为节点打标签

[ deployment]# kubectl label node 192.168.132.132 aaa=bbb

node/192.168.132.132 labeled

[ deployment]# kubectl get pods 

NAME                                READY   STATUS              RESTARTS   AGE
busybox-546555c84-2psbb             1/1     Running             13         24h
busybox-674bd96f74-m4fst            0/1     ContainerCreating   0          7m26s

[ deployment]# kubectl describe pods busybox-674bd96f74-m4fst

Events:
  Type     Reason            Age        From                      Message
  ----     ------            ----       ----                      -------
  Warning  FailedScheduling  <unknown>  default-scheduler         0/3 nodes are available: 3 node(s) didn‘t match node selector.
  Warning  FailedScheduling  <unknown>  default-scheduler         0/3 nodes are available: 3 node(s) didn‘t match node selector.
  Normal   Scheduled         <unknown>  default-scheduler         Successfully assigned default/busybox-674bd96f74-m4fst to 192.168.132.132
  Normal   Pulling           38s        kubelet, 192.168.132.132  Pulling image "busybox"
  Normal   Pulled            34s        kubelet, 192.168.132.132  Successfully pulled image "busybox"
  Normal   Created           34s        kubelet, 192.168.132.132  Created container busybox
  Normal   Started           33s        kubelet, 192.168.132.132  Started container busybox

2.5 删除标签

[ deployment]# kubectl label node 192.168.132.132  aaa-

删除标签后,已经运行的pod会继续运行

[ deployment]# kubectl get pods -o wide

NAME                                READY   STATUS    RESTARTS   AGE     IP            NODE              NOMINATED NODE   READINESS GATES
busybox-674bd96f74-m4fst            1/1     Running   0          12m     10.244.1.29   192.168.132.132   <none>           <none>
goproxy                             1/1     Running   1          3d11h   10.244.1.21   192.168.132.132   <none>           <none>

2.6 删除pods

[ deployment]# kubectl delete pods busybox-674bd96f74-m4fst

pod "busybox-674bd96f74-m4fst" deleted

deployment]# kubectl get pods -o wide

NAME                                READY   STATUS    RESTARTS   AGE     IP            NODE              NOMINATED NODE   READINESS GATES
busybox-674bd96f74-8d7ml            0/1     Pending   0          65s     <none>        <none>            <none>           <none>

一直不能启动,是因为没有匹配的标签

2.7 自定义标签

为192.168.132.132 打自定义标签

[ deployment]# kubectl label node 192.168.132.132  ingress=enable

node/192.168.132.132 labeled

[ deployment]# vim /yamls/ingress/nginx-controller.yaml

nodeSelector:
        ingress: enable

[ deployment]# kubectl apply -f /yamls/ingress/nginx-controller.yaml

namespace/ingress-nginx unchanged
configmap/nginx-configuration unchanged
configmap/tcp-services unchanged
configmap/udp-services unchanged
serviceaccount/nginx-ingress-serviceaccount unchanged
clusterrole.rbac.authorization.k8s.io/nginx-ingress-clusterrole unchanged
role.rbac.authorization.k8s.io/nginx-ingress-role unchanged
rolebinding.rbac.authorization.k8s.io/nginx-ingress-role-nisa-binding unchanged
clusterrolebinding.rbac.authorization.k8s.io/nginx-ingress-clusterrole-nisa-binding unchanged
deployment.apps/nginx-ingress-controller configured
limitrange/ingress-nginx configured

[ deployment]# kubectl get pods -n ingress-nginx -o wide

NAME                                        READY   STATUS        RESTARTS   AGE   IP                NODE              NOMINATED NODE   READINESS GATES
nginx-ingress-controller-5cffd956bf-dm9qf   1/1     Terminating   0          32m   192.168.132.133   192.168.132.133   <none>           <none>
nginx-ingress-controller-79669b846b-nlrxl   1/1     Running       0          16s   192.168.132.132   192.168.132.132   <none>           <none>

[ deployment]# kubectl get pods -n ingress-nginx -o wide

NAME                                        READY   STATUS    RESTARTS   AGE   IP                NODE              NOMINATED NODE   READINESS GATES
nginx-ingress-controller-79669b846b-nlrxl   1/1     Running   0          95s   192.168.132.132   192.168.132.132   <none>           <none>

另一种,是直接使用nodeName:指定节点IP

[ deployment]# vim /yamls/ingress/nginx-controller.yaml 

hostNetwork: true
      nodeName: "192.168.132.133"

[ deployment]# kubectl get nodes

NAME              STATUS   ROLES    AGE     VERSION
192.168.132.131   Ready    master   3d19h   v1.17.0
192.168.132.132   Ready    <none>   3d19h   v1.17.0
192.168.132.133   Ready    <none>   3d19h   v1.17.0

尽量不要指定nodeName,因为当这个节点断掉之后,必须修改新的nodeName,但是使用标签选择器,直接打标签就可以了

可以这么操作,运行节点的副本设为2,然后给其中两个节点打上相同的标签,在使用标签选择器选择标签,这样就可以让ingress运行在固定的节点上


博主声明:本文的内容来源主要来自誉天教育晏威老师,由本人实验完成操作验证,需要的博友请联系誉天教育(http://www.yutianedu.com/),获得官方同意或者晏老师(https://www.cnblogs.com/breezey/)本人同意即可转载,谢谢!

相关推荐