Kubernetes 集群安装部署

最新推荐文章于 2021-03-31 14:07:14 发布

weixin_30293135

最新推荐文章于 2021-03-31 14:07:14 发布

阅读量64

点赞数

文章标签：运维人工智能 json

原文链接：http://www.cnblogs.com/wangjq19920210/p/9262156.html

版权

etcd集群配置

master节点配置

1.安装kubernetes etcd

1 [root@k8s ~]# yum -y install kubernetes-master etcd

2.配置 etcd 选项

 1 [root@k8s ~]# cat /etc/etcd/etcd.conf 
 2 #[Member]
 3 ETCD_DATA_DIR="/var/lib/etcd/default.etcd"
 4 ETCD_LISTEN_PEER_URLS="http://172.19.15.92:2380"
 5 ETCD_LISTEN_CLIENT_URLS="http://172.19.15.92:2379,http://127.0.0.1:2379"
 6 ETCD_MAX_SNAPSHOTS="5"
 7 ETCD_NAME="etcd1"
 8 ETCD_HEARTBEAT_INTERVAL=6000     
 9 ETCD_ELECTION_TIMEOUT=30000
10 
11 #[Clustering]
12 ETCD_INITIAL_ADVERTISE_PEER_URLS="http://172.19.15.92:2380"
13 ETCD_ADVERTISE_CLIENT_URLS="http://172.19.15.92:2379"
14 ETCD_INITIAL_CLUSTER="etcd1=http://172.19.15.92:2380,etcd2=http://172.19.15.93:2380,etcd3=http://172.19.15.94:2380"

nodes节点配置

1.安装部署kubernetes-node /etcd /flannel /docker

1 [root@k8s-node1 ~]# yum -y install kubernetes-node etcd flannel docker

2.分别配置etcd，node1 与 node2 的配置方法相同，以 node1 配置文件为例说明

 1 [root@k8s-node1 ~]# cat /etc/etcd/etcd.conf 
 2 #[Member]
 3 ETCD_DATA_DIR="/var/lib/etcd/default.etcd"
 4 ETCD_LISTEN_PEER_URLS="http://172.19.15.93:2380"
 5 ETCD_LISTEN_CLIENT_URLS="http://172.19.15.93:2379,http://127.0.0.1:2379"
 6 ETCD_NAME="etcd2"
 7 ETCD_HEARTBEAT_INTERVAL=6000     
 8 ETCD_ELECTION_TIMEOUT=30000
 9 
10 #[Clustering]
11 ETCD_INITIAL_ADVERTISE_PEER_URLS="http://172.19.15.93:2380"
12 ETCD_ADVERTISE_CLIENT_URLS="http://172.19.15.93:2379"
13 ETCD_INITIAL_CLUSTER="etcd1=http://172.19.15.92:2380,etcd2=http://172.19.15.93:2380,etcd3=http://172.19.15.94:2380"

启动etcd cluster

分别在3台服务器启动etcd

 1 [root@k8s ~]# systemctl start etcd.service
 2 [root@k8s ~]# systemctl status etcd.service -l
 3 ● etcd.service - Etcd Server
 4    Loaded: loaded (/usr/lib/systemd/system/etcd.service; enabled; vendor preset: disabled)
 5    Active: active (running) since 二 2018-07-03 18:13:06 CST; 16h ago
 6  Main PID: 2085 (etcd)
 7     Tasks: 31
 8    Memory: 328.4M
 9    CGroup: /system.slice/etcd.service
10            └─2085 /usr/bin/etcd --name=etcd1 --data-dir=/var/lib/etcd/default.etcd --listen-client-urls=http://172.19.15.92:2379,http://127.0.0.1:2379

查看etcd集群状态

1 [root@k8s ~]# etcdctl cluster-health
2 member 8c24796af2c20350 is healthy: got healthy result from http://172.19.15.94:2379
3 member e66597512233d97d is healthy: got healthy result from http://172.19.15.93:2379
4 member edfc36869b54e803 is healthy: got healthy result from http://172.19.15.92:2379
5 cluster is healthy

Kubernetes集群配置

master节点配置

1.apiserver配置文件修改，注意KUBE_ADMISSION_CONTROL选项的参数配置

1 [root@k8s ~]# cat /etc/kubernetes/apiserver
2 KUBE_API_ADDRESS="--insecure-bind-address=0.0.0.0"
3 KUBE_API_PORT="--port=8080"
4 KUBELET_PORT="--kubelet-port=10250"
5 KUBE_ETCD_SERVERS="--etcd-servers=http://172.19.15.92:2379,http://172.19.15.93:2379,http://172.19.15.94:2379"
6 KUBE_SERVICE_ADDRESSES="--service-cluster-ip-range=10.254.0.0/16"
7 KUBE_ADMISSION_CONTROL="--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota"
8 KUBE_API_ARGS=""

2.启动服务

1 [root@k8s ~]# systemctl start kube-apiserver
2 [root@k8s ~]# systemctl start kube-controller-manager
3 [root@k8s ~]# systemctl start kube-scheduler
4 [root@k8s ~]# systemctl enable kube-apiserver
5 [root@k8s ~]# systemctl enable kube-controller-manager
6 [root@k8s ~]# systemctl enable kube-scheduler

nodes节点配置

1.配置config配置，node1&node2配置相同，以node1为例说明

1 [root@k8s-node1 ~]# cat /etc/kubernetes/config
2 KUBE_LOGTOSTDERR="--logtostderr=true"
3 KUBE_LOG_LEVEL="--v=0"
4 KUBE_ALLOW_PRIV="--allow-privileged=false"
5 KUBE_MASTER="--master=http://172.19.15.92:8080"

2.配置kubelet

1 [root@k8s-node1 ~]# cat /etc/kubernetes/kubelet 
2 KUBELET_ADDRESS="--address=127.0.0.1"
3 KUBELET_HOSTNAME="--hostname-override=172.19.15.93"
4 KUBELET_API_SERVER="--api-servers=http://172.19.15.92:8080"
5 KUBELET_POD_INFRA_CONTAINER="--pod-infra-container-image=registry.access.redhat.com/rhel7/pod-infrastructure:latest"
6 KUBELET_ARGS=""

3.docker service配置文件

 1 [root@k8s-node1 ~]# cat /usr/lib/systemd/system/docker.service
 2 [Unit]
 3 Description=Docker Application Container Engine
 4 Documentation=https://docs.docker.com
 5 After=network.target firewalld.service
 6 
 7 [Service]
 8 Type=notify
 9 Environment="http_proxy=http://192.168.59.241:8888/" "https_proxy=https://192.168.59.241:8888/"
10 ExecStart=/usr/bin/dockerd --registry-mirror=http://f2d6cb40.m.daocloud.io --bip=192.100.90.1/24
11 ExecReload=/bin/kill -s HUP $MAINPID
12 LimitNOFILE=infinity
13 LimitNPROC=infinity
14 LimitCORE=infinity
15 TimeoutStartSec=0
16 Delegate=yes
17 KillMode=process
18 
19 [Install]
20 WantedBy=multi-user.target

网络配置

这里使用flannel进行网络配置，已经在2个节点上安装，下面进行配置。

在节点上进行配置flannel

1 [root@k8s-node1 ~]# cat /etc/sysconfig/flanneld
2 FLANNEL_ETCD_ENDPOINTS="http://172.19.15.92:2379,http://172.19.15.93:2379,http://172.19.15.94:2379"
3 FLANNEL_ETCD_PREFIX="/k8s/network"
4 FLANNEL_OPTIONS="--logtostderr=true --log_dir=/var/log/k8s/flannel/ --etcd-prefix=/k8s/network  --etcd-endpoints=http://172.19.15.92:2379,http://172.19.15.93:2379,http://172.19.15.94:2379 --iface=ens160"

master 节点需要配置 etcd 网络：

1 etcdctl set /k8s/network/config '{"Network":"192.100.0.1/16"}'

2.启动服务

1 [root@k8s-node1 ~]# systemctl start kubelet
2 [root@k8s-node1 ~]# systemctl start docker
3 [root@k8s-node1 ~]# systemctl start flanneld
4 [root@k8s-node1 ~]# systemctl enable kubelet
5 [root@k8s-node1 ~]# systemctl enable docker
6 [root@k8s-node1 ~]# systemctl enable flanneld

查看集群状态

 1 [root@k8s ~]# kubectl get nodes
 2 NAME           STATUS    AGE
 3 172.19.15.92   Ready     16h
 4 172.19.15.93   Ready     1d
 5 172.19.15.94   Ready     1d
 6 [root@k8s ~]# etcdctl  member list
 7 8c24796af2c20350: name=etcd3 peerURLs=http://172.19.15.94:2380 clientURLs=http://172.19.15.94:2379 isLeader=false
 8 e66597512233d97d: name=etcd2 peerURLs=http://172.19.15.93:2380 clientURLs=http://172.19.15.93:2379 isLeader=false
 9 edfc36869b54e803: name=etcd1 peerURLs=http://172.19.15.92:2380 clientURLs=http://172.19.15.92:2379 isLeader=true
10 [root@k8s ~]# etcdctl cluster-health
11 member 8c24796af2c20350 is healthy: got healthy result from http://172.19.15.94:2379
12 member e66597512233d97d is healthy: got healthy result from http://172.19.15.93:2379
13 member edfc36869b54e803 is healthy: got healthy result from http://172.19.15.92:2379
14 cluster is healthy

更改 docker 网段为 flannel 分配的网段

1 # export FLANNEL_SUBNET=10.254.26.1/24
2 # cat << EOF > /etc/docker/daemon.json
3 {
4   "bip" : "$FLANNEL_SUBNET"
5 }
6 EOF
7 # systemctl daemon-reload
8 # systemctl restart docker