這個專案(http://github.com/kubernetes/kube-deploy) 大概是kubernetes的安裝上數一數二簡單的方式... 只要多玩幾次，應該可以很快上手唷～
首先建立cluster用的主機，並安裝docker...(這邊不再贅述安裝過程)，目前測試過成功的安裝包括：
- GCE CentOS7 (docker 1.12.0)
- GCE CoreOS stable (docker 1.10)
- GCE Ubuntu 15 (docker 1.12.0)
以官方的說法，應該docker 1.10以上版本就可以安裝... 其中CentOS7可以完全不修改的情況下照著流程跑完，其他的環境可以修改一小部分安裝腳本來跑完...
進到master主機，執行下面指令 ( 確認是 root 權限 )
git clone https://github.com/kubernetes/kube-deploy
cd kube-deploy/docker-multinode
bash ./master.sh
進到worker主機，執行下面指令 ( 確認是 root 權限 )
git clone https://github.com/kubernetes/kube-deploy
cd kube-deploy/docker-multinode
export MASTER_IP=[master_ip]
bash ./worker.sh
最後啟動完，在master與worker上應該會看到如下的程序列表...
docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
b5abf6fa4452 gcr.io/google_containers/exechealthz-amd64:1.1 "/exechealthz '-cmd=n" 59 minutes ago Up 59 minutes k8s_healthz.92b64eae_kube-dns-v17.1-0bxw2_kube-system_11900aae-5fa4-11e6-b1a0-42010af00009_d81566f1
5f175ce7e83b gcr.io/google_containers/kube-dnsmasq-amd64:1.3 "/usr/sbin/dnsmasq --" 59 minutes ago Up 59 minutes k8s_dnsmasq.9cc9760b_kube-dns-v17.1-0bxw2_kube-system_11900aae-5fa4-11e6-b1a0-42010af00009_b8c0d776
35fcdba200d8 gcr.io/google_containers/kubedns-amd64:1.5 "/kube-dns --domain=c" About an hour ago Up 59 minutes k8s_kubedns.6b818f53_kube-dns-v17.1-0bxw2_kube-system_11900aae-5fa4-11e6-b1a0-42010af00009_4259f590
0212754549a0 gcr.io/google_containers/kubernetes-dashboard-amd64:v1.1.0 "/dashboard --port=90" About an hour ago Up About an hour k8s_kubernetes-dashboard.f96ccb63_kubernetes-dashboard-v1.1.0-7wupe_kube-system_118a6eba-5fa4-11e6-b1a0-42010af00009_8c6719bc
e93ee4079ab1 gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD.a6b39ba7_kube-dns-v17.1-0bxw2_kube-system_11900aae-5fa4-11e6-b1a0-42010af00009_9b594f6f
c131cd01bc56 gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD.2225036b_kubernetes-dashboard-v1.1.0-7wupe_kube-system_118a6eba-5fa4-11e6-b1a0-42010af00009_9d1b3989
3a22fbb92e12 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/copy-addons.sh" About an hour ago Up About an hour k8s_kube-addon-manager-data.3b72199f_kube-addon-manager-10.240.0.9_kube-system_29f6ff0d8867cc4a121f6442895ecd07_f321e682
3ef310f4d228 gcr.io/google-containers/kube-addon-manager-amd64:v4 "/opt/kube-addons.sh" About an hour ago Up About an hour k8s_kube-addon-manager.a97f8ca4_kube-addon-manager-10.240.0.9_kube-system_29f6ff0d8867cc4a121f6442895ecd07_3db38dd0
ccdd4998d981 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube scheduler" About an hour ago Up About an hour k8s_scheduler.33afcc1_k8s-master-10.240.0.9_kube-system_c25065f253ce380787b6c4a8ca9badb5_9aad247d
807dd8ff0b84 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube apiserver" About an hour ago Up About an hour k8s_apiserver.e1a5bf74_k8s-master-10.240.0.9_kube-system_c25065f253ce380787b6c4a8ca9badb5_6aa3f724
27429ab818be gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube controlle" About an hour ago Up About an hour k8s_controller-manager.72ff5161_k8s-master-10.240.0.9_kube-system_c25065f253ce380787b6c4a8ca9badb5_03eb94d9
b3655efaf6ac gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/setup-files.sh IP:1" About an hour ago Up About an hour k8s_setup.39693812_k8s-master-10.240.0.9_kube-system_c25065f253ce380787b6c4a8ca9badb5_c740fd19
d346fa48600f gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD.d8dbe16c_k8s-master-10.240.0.9_kube-system_c25065f253ce380787b6c4a8ca9badb5_e3014663
3a574093cc0c gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD.d8dbe16c_kube-addon-manager-10.240.0.9_kube-system_29f6ff0d8867cc4a121f6442895ecd07_91be59bb
1b5a60af23b1 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube proxy --m" About an hour ago Up About an hour k8s_kube-proxy.a4344856_k8s-proxy-10.240.0.9_kube-system_d03a8e5b4dd1dbcc1ae1a06eb6c88afb_b596bc54
563acb2f8f5b gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD.d8dbe16c_k8s-proxy-10.240.0.9_kube-system_d03a8e5b4dd1dbcc1ae1a06eb6c88afb_587709fd
3b56a88f302f gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube kubelet -" About an hour ago Up About an hour kube_kubelet_80b03
docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
f812c56ec7e7 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube proxy --m" 5 minutes ago Up 4 minutes kube_proxy_acf16
cd9b2387c081 gcr.io/google_containers/hyperkube-amd64:v1.3.4 "/hyperkube kubelet -" 5 minutes ago Up 5 minutes kube_kubelet_a5ddd
如果遇到執行master.sh或slave.sh時候，有/var/lib/kubelet不是shared mount filesystem時候，可以透過下面方式解決
Solution 1: 將該mount point改為shared mount...
DOCKER_CONF=$(systemctl cat docker | head -1 | awk '{print $2}')
sed -i.bak 's/^\(MountFlags=\).*/\1shared/' $DOCKER_CONF
systemctl daemon-reload
systemctl restart docker
mkdir -p /var/lib/kubelet
mount --bind /var/lib/kubelet /var/lib/kubelet
mount --make-shared /var/lib/kubelet
Solution 2: 直接移除啟動腳本的shared flag
vi kube-deploy/docker-multinode/common.sh
修改mount /var/lib/kubelet那段的shared flag
...
-v /var/lib/kubelet:/var/lib/kubelet:shared \
...
修改為：
...
-v /var/lib/kubelet:/var/lib/kubelet \
...
可能是在某些環境的docker啟動時間不太一定，曾經在ubuntu與coreos上遇到最後docker run會有docker為執行的錯誤...後來透過在啟動前停止3秒鐘來避免找不到環境的問題...
...
kube::multinode::start_k8s_master() {
kube::log::status "Launching Kubernetes master components..."
kube::multinode::make_shared_kubelet_dir
# Stop 3 seconds
sleep 3
docker run -d \
--net=host \
--pid=host \
--privileged \
--restart=${RESTART_POLICY} \
--name kube_kubelet_$(kube::helpers::small_sha) \
...
如果遇到k8s服務需要重新啟動的時候，如果要保留上面執行的服務... 需要在啟動時候注意一下別把/var/lib/kubelet的目錄清空了...
[root@simon-k8s-master docker-multinode]# ./master.sh
+++ [0811 10:37:04] K8S_VERSION is set to: v1.3.4
+++ [0811 10:37:04] ETCD_VERSION is set to: 2.2.5
+++ [0811 10:37:04] FLANNEL_VERSION is set to: 0.5.5
+++ [0811 10:37:04] FLANNEL_IPMASQ is set to: true
+++ [0811 10:37:04] FLANNEL_NETWORK is set to: 10.1.0.0/16
+++ [0811 10:37:04] FLANNEL_BACKEND is set to: udp
+++ [0811 10:37:04] RESTART_POLICY is set to: unless-stopped
+++ [0811 10:37:04] MASTER_IP is set to: localhost
+++ [0811 10:37:04] ARCH is set to: amd64
+++ [0811 10:37:04] IP_ADDRESS is set to: 10.240.0.9
+++ [0811 10:37:04] USE_CNI is set to: false
+++ [0811 10:37:04] --------------------------------------------
+++ [0811 10:37:04] Killing all kubernetes containers...
dd34871b72fe
8387d65e2963
e7ac535047fa
5e70475c9409
b644a3a7c4b5
711df2fabb08
c0f45e19fa1b
5d68320c8c2f
cdc7e16bad99
Do you want to clean /var/lib/kubelet? [Y/n] n
...(skip)
- 關於shared mount: https://linux.cn/article-5057-5.html
- 官方multi-node安裝說明文件：http://kubernetes.io/docs/user-guide/pods/multi-container/
- Github: https://github.com/kubernetes/kube-deploy