2018/1/4

配置 worker 節點

• 初始化
• 加入集群
• 切換 worker 節點連接到 apiserver 的 LB 入口
• 調整集群中節點角色和調度策略

初始化

• /etc/hosts

  
  ### k8s master @envDev
  10.10.9.67 tvm-00
  10.10.9.68 tvm-01
  10.10.9.69 tvm-02

k8s worker @envDev

10.10.9.74 tvm-03
10.10.9.75 tvm-04

k8s apiserver SLB

10.10.9.76 kubernetes.default.svc.cluster.local

- docker
```bash
yum -y install yum-utils
yum-config-manager --add-repo http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
yum makecache fast
yum -y install docker-ce-17.09.1.ce-1.el7.centos.x86_64
systemctl enable docker

mkdir -p /data2/docker
mkdir -p /etc/docker
tee /etc/docker/daemon.json <<-‘EOF‘
{
  "exec-opts": ["native.cgroupdriver=cgroupfs"],
  "graph": "/data2/docker",
  "storage-driver": "overlay",
  "log-driver": "json-file",
  "log-opts": {
    "max-size": "100m"
  },
  "registry-mirrors": ["https://xxx.mirror.aliyuncs.com"]
}
EOF
 

• k8s

  
  cat <<‘_EOF‘ >  /etc/sysctl.d/k8s.conf
  net.bridge.bridge-nf-call-ip6tables = 1
  net.bridge.bridge-nf-call-iptables = 1
  _EOF
  sysctl --system

worker 節點需要的鏡像

docker load -i gcr.io-worker.tar
docker load -i calico-worker.tar

worker 節點需要的 rpm 包

[root@tvm-03 ~]# cd ~/k8s_install/k8s_rpms_1.9
[root@tvm-03 k8s_rpms_1.9]# ls

yum localinstall *.rpm -y
sed -i ‘s#--cgroup-driver=systemd#--cgroup-driver=cgroupfs#‘ /etc/systemd/system/kubelet.service.d/10-kubeadm.conf
systemctl daemon-reload

##### 加入集群
```bash
### 在 mater 上重新生成 token （如果過期了的話）
kubeadm join --token 9db9dd.09bd5226fb3f259c 10.10.9.67:6443 --discovery-token-ca-cert-hash sha256:42cfdc412e731793ce2fa20aad1d8163ee8e6e5c05c30765f204ff086823c653

### 在 master 上查看節點是否加入
[root@tvm-00 ~]# kubectl get nodes
NAME                     STATUS    ROLES     AGE       VERSION
tvm-00   Ready     master    12d       v1.9.0
tvm-01   Ready     <none>    8d        v1.9.0
tvm-02   Ready     <none>    8d        v1.9.0
tvm-03   Ready     <none>    30m       v1.9.0
tvm-04   Ready     <none>    29m       v1.9.0

切換 worker 節點連接到 apiserver 的 LB 入口

sed -i ‘s#https://10.10.9.67:6443#https://kubernetes.default.svc.cluster.local:6443#‘ /etc/kubernetes/kubelet.conf
systemctl restart kubelet

### 驗證一下
[root@tvm-03 k8s_install]# ss -antp4 |grep ‘:6443‘
ESTAB      0      0      10.10.9.74:56118              10.10.9.76:6443                users:(("kubelet",pid=21260,fd=12))
ESTAB      0      0      10.10.9.74:53204              10.10.9.67:6443                users:(("kube-proxy",pid=15834,fd=5))

### 符合預期

調整集群中節點角色和調度策略

### 當前狀態
[root@tvm-00 ~]# kubectl describe nodes/tvm-00 |grep -E ‘(Roles|Taints)‘
Roles:              master
Taints:             node-role.kubernetes.io/master:NoSchedule
[root@tvm-00 ~]# kubectl describe nodes/tvm-01 |grep -E ‘(Roles|Taints)‘
Roles:              <none>
Taints:             <none>

### 設置 tvm-01 為 master 且不允許調度
[root@tvm-00 ~]# kubectl label node tvm-01 node-role.kubernetes.io/master=
node "tvm-01" labeled
[root@tvm-00 ~]# kubectl taint nodes tvm-01 node-role.kubernetes.io/master=:NoSchedule
node "tvm-01" tainted

### 符合預期
[root@tvm-00 ~]# kubectl describe nodes/tvm-01 |grep -E ‘(Roles|Taints)‘
Roles:              master
Taints:             node-role.kubernetes.io/master:NoSchedule

### 設置 tvm-02 為 master 且不允許調度（操作類似，略）
[root@tvm-00 ~]# kubectl get nodes
NAME                     STATUS    ROLES     AGE       VERSION
tvm-00   Ready     master    12d       v1.9.0
tvm-01   Ready     master    8d        v1.9.0
tvm-02   Ready     master    8d        v1.9.0
tvm-03   Ready     <none>    1h        v1.9.0
tvm-04   Ready     <none>    1h        v1.9.0

ZYXW、參考

1. Taints and Tolerations

使用kubeadm部署k8s集群09-配置worker節點