江湖有缘

江湖有缘

关注
发信
关注(28)粉丝(399)

kubernetes集群之调度系统

kubernetes集群之调度系统

一、kube-scheduler介绍

1.kube-scheduler简介

2.k8s的调度系统作用

3.kubernetes组件示意图

kubernetes集群之调度系统-LMLPHP

4.schedule调度器工作示意图

kubernetes集群之调度系统-LMLPHP

二、查看kubernetes状态

[root@k8s-master ~]# kubectl get nodes -owide
NAME         STATUS   ROLES                  AGE   VERSION   INTERNAL-IP     EXTERNAL-IP   OS-IMAGE                KERNEL-VERSION          CONTAINER-RUNTIME
k8s-master   Ready    control-plane,master   39h   v1.23.1   192.168.3.201   <none>        CentOS Linux 7 (Core)   3.10.0-957.el7.x86_64   containerd://1.6.6
k8s-node01   Ready    worker                 39h   v1.23.1   192.168.3.202   <none>        CentOS Linux 7 (Core)   3.10.0-957.el7.x86_64   containerd://1.6.6
k8s-node02   Ready    <none>                 39h   v1.23.1   192.168.3.203   <none>        CentOS Linux 7 (Core)   3.10.0-957.el7.x86_64   containerd://1.6.6

三、kube-scheduler的选择节点流程

1.预选

#一个pod中容器运行的资源按要求
resources:
    request:
       cpu: 1
       memory: 1Gi

2.优选

打分的参考项
1.节点的实际资源占用
2.节点中pod的个数
3.节点中的cpu负载情况
4.节点中内存的使用情况
.......

3.终选

四、干预调度方法-标签选择器

1.标签选择器介绍

2.标签选择器的yaml写法

①查看所有节点标签

[root@k8s-master ~]# kubectl get nodes --show-labels
NAME         STATUS   ROLES                  AGE   VERSION   LABELS
k8s-master   Ready    control-plane,master   40h   v1.23.1   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-master,kubernetes.io/os=linux,node-role.kubernetes.io/control-plane=,node-role.kubernetes.io/master=,node.kubernetes.io/exclude-from-external-load-balancers=
k8s-node01   Ready    worker                 40h   v1.23.1   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-node01,kubernetes.io/os=linux,node-role.kubernetes.io/worker=
k8s-node02   Ready    <none>                 40h   v1.23.1   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=k8s-node02,kubernetes.io/os=linux,type=dell730

②在yaml文件中选择标签

volumes:
- name: rootdir
  hostPath:
      path: /data/mysql
nodeSelector:
  #disk: ssd
  kubernetes.io/hostname=k8s-node01  #可以选择系统内置的独一无二的标签
containers:
- name: mysql
  image: mysql:5.7
  volumeMounts:
  - name: rootdir
    mountPath: /var/lib/mysql

3.运行一个完整pod示例

apiVersion: apps/v1
kind: Deployment
metadata:
  creationTimestamp: null
  labels:
    app: mysql
  name: mysql
spec:
  replicas: 1
  selector:
    matchLabels:
      app: mysql
  strategy: {}
  template:
    metadata:
      creationTimestamp: null
      labels:
        app: mysql
    spec:
      volumes:
      - name: datadir
        hostPath:
           path: /data/mysql
      nodeSelector:
        disk: ssd
        nodeType: cpu
      containers:
      - image: mysql:5.7
        name: mysql
        env:
        - name: MYSQL_ROOT_PASSWORD
          value: "redhat"
        volumeMounts:
        - name: datadir
          mountPath: /var/lib/mysql

4.给node02节点添加标签

[root@k8s-master ~]# kubectl label nodes k8s-node02 disk=ssd
node/k8s-node02 labeled
[root@k8s-master ~]# kubectl label nodes k8s-node02 nodeType=cpu
node/k8s-node02 labeled

5.创建pod

[root@k8s-master ~]# kubectl apply -f ./label.yaml
deployment.apps/mysql created

6.查看pod所在节点

[root@k8s-master ~]# kubectl get pod -owide |grep node02
elasticsearch-master-0                    1/1     Running   2 (7m54s ago)    29h   10.244.58.224   k8s-node02   <none>           <none>
fb-filebeat-lj5p7                         1/1     Running   3 (7m54s ago)    28h   10.244.58.221   k8s-node02   <none>           <none>
kb-kibana-5c46dbc5dd-htw7n                1/1     Running   1 (7m54s ago)    25h   10.244.58.222   k8s-node02   <none>           <none>
metric-metricbeat-5h5g5                   1/1     Running   2 (7m53s ago)    27h   192.168.3.203   k8s-node02   <none>           <none>
metric-metricbeat-758c5c674-ldgg4         1/1     Running   2 (7m54s ago)    27h   10.244.58.225   k8s-node02   <none>           <none>
mysql-59c6fc696d-qrjx9                    1/1     Running   0                13m   10.244.58.223   k8s-node02   <none>           <none>

五、干预调度方法——污点

1.污点taint介绍

2.污点类型

* preferNoSchedule：
尽可能的不调度

* NoSchedule： 不调度
当前node如果打伤污点之前已经有一些pod运行正在上面，当打上污点后，新的pod不会调度其上；但是已经运行的pod不会被驱逐

* NoExecute： 不调度 .
当前node如果打上污点之前已经有一些pod运行正在上面，当打上污点后，会立即驱逐现有pod

3.为工作节点创建污点

[root@k8s-master ~]# kubectl taint node k8s-node02 key1=value:NoSchedule
node/k8s-node02 tainted
[root@k8s-master ~]# kubectl taint node k8s-node02 key2=value:NoExecute
node/k8s-node02 tainted

4.删除工作节点上污点

kuberctl taint node k8s-node02   key1-

5.查看某个节点的污点

[root@k8s-master ~]# kubectl describe nodes k8s-node02 |grep -i tain -A2 -B2
                    nodeType=cpu
                    type=dell730
Annotations:        kubeadm.alpha.kubernetes.io/cri-socket: /run/containerd/containerd.sock
                    node.alpha.kubernetes.io/ttl: 0
                    projectcalico.org/IPv4Address: 192.168.3.203/24
--
                    volumes.kubernetes.io/controller-managed-attach-detach: true
CreationTimestamp:  Sun, 03 Jul 2022 01:14:11 +0800
Taints:             key2=value:NoExecute
                    key1=value:NoSchedule
Unschedulable:      false
--
  Operating System:           linux
  Architecture:               amd64
  Container Runtime Version:  containerd://1.6.6
  Kubelet Version:            v1.23.1
  Kube-Proxy Version:         v1.23.1

六、干预调度方法——容忍

1.容忍toerations介绍

2.容忍和污点的pod选择

kubernetes集群之调度系统-LMLPHP

3.在yaml文件中容忍用法

apiVersion: v1
kind: Pod
metadata:
  name: nginx
  labels:
    app: nginx
spec:
  containers:
  - name: nginx
    image: nginx:1.5.2
  tolerations:
  - key: "check"
    operator: "Equal"
    value: "xtaint"
    effect: "NoExecute"
    tolerationSeconds: 3600

4.容忍相关参数解释

tolerations:----------->容忍
- key: "check" ----------->容忍的键
  operator: "Equal"----------->操作符"等于"
  value: "xtaint"----------->容忍的键对应的键值
  effect: "NoExecute"----------->容忍的键对应的影响效果
  tolerationSeconds: 3600----------->容忍3600秒。这个pod也不会像普通pod那样立即被驱逐，而是再等上3600秒才被删除。
07-05 07:28
Powered by LMLPHP ©2024 江湖有缘 0.051892