centos7.4安裝配置使用簡介Kubernetes (k8s)

Kubernetes集羣部署

架構說明

Kubernetes集羣組件:

• etcd 一個高可用的K/V鍵值對存儲和服務發現系統
• flannel 實現誇主機的容器網絡的通訊
• kube-apiserver 提供kubernetes集羣的API調用
• kube-controller-manager 確保集羣服務
• kube-scheduler 調度容器，分配到Node
• kubelet 在Node節點上按照配置文件中定義的容器規格啓動容器
• kube-proxy 提供網絡代理服務

節點	IP地址
master	10.10.10.14
node1	10.10.10.15
node2	10.10.10.16

更改Hostname爲 master、node一、node2，配置全部測試機的/etc/hosts文件

[root@master ~]# cat /etc/hosts
10.10.10.14 master  etcd  node14
10.10.10.15 node1   node15
10.10.10.16 node2   node16

關閉CentOS7自帶的防火牆服務

安裝

系統初始化安裝（全部主機）-選擇【最小化安裝】,而後yum update,升級到最新版本
yum -y install epel-release
yum update

[root@master ~]#

yum install -y etcd kubernetes-master ntp flannel

[root@node1 ~]#

yum install -y kubernetes-node ntp flannel docker

時間校對
全部主機

systemctl start ntpd;systemctl enable ntpd
ntpdate ntp1.aliyun.com
hwclock -w

配置etcd服務器

[root@master ~]# grep -v '^#' /etc/etcd/etcd.conf

ETCD_NAME=default
ETCD_DATA_DIR="/var/lib/etcd/default.etcd"
ETCD_LISTEN_CLIENT_URLS="http://localhost:2379,http://10.10.10.14:2379"
ETCD_ADVERTISE_CLIENT_URLS="http://10.10.10.14:2379"

啓動服務

systemctl start etcd;systemctl enable etcd

檢查etcd cluster狀態

[root@master ~]# etcdctl cluster-health   
member 8e9e05c52164694d is healthy: got healthy result from http://10.10.10.14:2379
cluster is healthy

檢查etcd集羣成員列表，這裏只有一臺

[root@master ~]# etcdctl member list
8e9e05c52164694d: name=default peerURLs=http://localhost:2380 clientURLs=http://10.10.10.14:2379 isLeader=true

配置master服務器

1) 配置kube-apiserver配置文件

[root@master ~]# grep -v '^#' /etc/kubernetes/config

KUBE_LOGTOSTDERR="--logtostderr=true"
KUBE_LOG_LEVEL="--v=0"
KUBE_ALLOW_PRIV="--allow-privileged=false"
KUBE_MASTER="--master=http://10.10.10.14:8080"

[root@master ~]# grep -v '^#' /etc/kubernetes/apiserver

KUBE_API_ADDRESS="--insecure-bind-address=0.0.0.0"
KUBE_ETCD_SERVERS="--etcd-servers=http://10.10.10.14:2379"
KUBE_SERVICE_ADDRESSES="--service-cluster-ip-range=10.254.0.0/16"
KUBE_ADMISSION_CONTROL="--admission-control=AlwaysAdmit"
KUBE_API_ARGS=""

2) 配置kube-controller-manager配置文件

[root@master ~]# grep -v '^#' /etc/kubernetes/controller-manager

KUBE_CONTROLLER_MANAGER_ARGS=""

3) 配置kube-scheduler配置文件

[root@master ~]# grep -v '^#' /etc/kubernetes/scheduler

KUBE_SCHEDULER_ARGS="--address=0.0.0.0"

4) 啓動服務

for i in  kube-apiserver kube-controller-manager kube-scheduler;do systemctl restart $i; systemctl enable $i;done

配置node1節點服務器

1) 配置etcd

[root@master ~]# etcdctl set /atomic.io/network/config '{"Network": "172.16.0.0/16"}'
{"Network": "172.16.0.0/16"}

2) 配置node1網絡，本實例採用flannel方式來配置，如需其餘方式，請參考Kubernetes官網。

[root@node1 ~]# grep -v '^#' /etc/sysconfig/flanneld 
FLANNEL_ETCD_ENDPOINTS="http://10.10.10.14:2379"
FLANNEL_ETCD_PREFIX="/atomic.io/network"
FLANNEL_OPTIONS=""

查看驗證網絡信息

[root@master ~]# etcdctl get /atomic.io/network/config  
{ "Network": "172.16.0.0/16" }
[root@master ~]# etcdctl ls /atomic.io/network/subnets
/atomic.io/network/subnets/172.16.69.0-24
/atomic.io/network/subnets/172.16.6.0-24
[root@master ~]# etcdctl get /atomic.io/network/subnets/172.16.6.0-24
{"PublicIP":"10.10.10.15"}
[root@master ~]# etcdctl get  /atomic.io/network/subnets/172.16.69.0-24
{"PublicIP":"10.10.10.16"}

3) 配置node1 kube-proxy

[root@node1 ~]# grep -v '^#' /etc/kubernetes/config 
KUBE_LOGTOSTDERR="--logtostderr=true"
KUBE_LOG_LEVEL="--v=0"
KUBE_ALLOW_PRIV="--allow-privileged=false"
KUBE_MASTER="--master=http://10.10.10.14:8080"

[root@node1 ~]# grep -v '^#' /etc/kubernetes/proxy                  
KUBE_PROXY_ARGS="--bind=address=0.0.0.0"
[root@node1 ~]#

4) 配置node1 kubelet

[root@node1 ~]# grep -v '^#' /etc/kubernetes/kubelet 
KUBELET_ADDRESS="--address=127.0.0.1"
KUBELET_HOSTNAME="--hostname-override=10.10.10.15"
KUBELET_API_SERVER="--api-servers=http://10.10.10.14:8080"
KUBELET_POD_INFRA_CONTAINER="--pod-infra-container-image=registry.access.redhat.com/rhel7/pod-infrastructure:latest"
KUBELET_ARGS=""

5) 啓動node1服務

for i in flanneld kube-proxy kubelet docker;do systemctl restart $i;systemctl enable $i;systemctl status $i ;done

配置node2節點服務器

node2與node1配置基本一致，除下面一處例外

[root@node2 ~]# vi /etc/kubernetes/kubelet
KUBELET_HOSTNAME="--hostname-override=10.10.10.16"

查看節點

[root@master ~]# kubectl get nodes   
NAME          STATUS    AGE
10.10.10.15   Ready     18h
10.10.10.16   Ready     13h

k8s支持2種方式,一種是直接經過命令參數的方式,另外一種是經過配置文件的方式,配置文件的話支持json和yaml

命令方式

創建pod

kubectl run nginx --image=nginx --port=80  --replicas=2

遇到問題

建立成功可是kubectl get pods 沒有結果
提示信息：no API token found for service account default
解決辦法：編輯/etc/kubernetes/apiserver 去除 KUBE_ADMISSION_CONTROL中的SecurityContextDeny,ServiceAccount，並重啓kube-apiserver.service服務

pod-infrastructure:latest鏡像下載失敗
報錯信息：image pull failed for registry.access.redhat.com/rhel7/pod-infrastructure:latest, this may be because there are no credentials on this request.
解決方案：yum install *rhsm* -y

登錄容器報錯
[root@node14 ~]# kubectl exec -it nginx-bl7lc /bin/bash
Error from server: error dialing backend: dial tcp 10.10.10.16:10250: getsockopt: connection refused
解決方法：
10250是kubelet的端口。
在Node上檢查/etc/kubernetes/kubelet
KUBELET_ADDRESS須要修改成node ip

命令查看

[root@master log]# kubectl get pods
NAME                     READY     STATUS    RESTARTS   AGE
nginx-3449338310-h6l9d   1/1       Running   0          6m
nginx-3449338310-n4grl   1/1       Running   0          6m

[root@master log]# kubectl get deployment
NAME      DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
nginx          2          2         2       2       13m

經過端口將應用鏈接到公網

[root@node14 log]# kubectl expose deployment nginx --port=80 --type=LoadBalancer
service "nginx" exposed

expose命令將會建立一個service，將本地（某個節點上）的一個隨機端口關聯到容器中的80端口。
可使用如下命令來查service：

[root@node14 log]# kubectl get service
NAME         CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
kubernetes   10.254.0.1       <none>        443/TCP        21h
nginx        10.254.160.227   <pending>     80:30255/TCP   7s

外網訪問：

[root@node16 ~]# elinks --dump http://10.10.10.16:30255
                           Welcome to nginx!

刪除deployment 與service

[root@node14 log]# kubectl delete deployment nginx
deployment "nginx" deleted
[root@node14 log]# kubectl delete service nginx
service "nginx" deleted

配置文件方式

定義pod 文件

[root@node14 ~]# vim nginx-pod.yaml

apiVersion: v1
kind: Pod
metadata:
  name: nginx
  labels:
     app: nginx
spec:
     containers:
     - name: nginx
       image: nginx
       imagePullPolicy: IfNotPresent
       ports:
       - containerPort: 80
     restartPolicy: Always

發佈到kubernetes集羣中

[root@node14 ~]# kubectl create -f nginx-pod.yaml 
pod "nginx" created

查看pod

[root@node14 ~]# kubectl get pods
NAME      READY     STATUS    RESTARTS   AGE
nginx     1/1       Running   0          16s

定義與之關聯的service 文件

[root@node14 ~]# vim nginx-svc.yaml

apiVersion: v1
kind: Service
metadata:
  name: nginx-service
spec:
  type: NodePort
  sessionAffinity: ClientIP
  selector:
    app: nginx
  ports:
    - port: 80
      nodePort: 30080

建立service

[root@node14 ~]# kubectl create -f nginx-svc.yaml 
service "nginx-service" created

查看剛剛建立的service

[root@node14 ~]# kubectl get service

NAME            CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
kubernetes      10.254.0.1       <none>        443/TCP        23h
nginx-service   10.254.154.111   <nodes>       80:30080/TCP   20s

驗證結果以下

[root@node16 log]# elinks --dump http://10.10.10.16:30080
                               Welcome to nginx!