前言
这是和我一步步部署kubernetes集群项目((fork自opsnull))中的一篇文章，下文是结合我之前部署kubernetes的过程产生的kuberentes环境，部署master节点的kube-apiserver、kube-controller-manager和kube-scheduler的过程。
高可用Kubernetes Master节点安装
kubernetes master 节点包含的组件：
- kube-apiserver
- kube-scheduler
- kube-controller-manager
目前这三个组件需要部署在同一台机器上。
kube-scheduler、kube-controller-manager和kube-apiserver三者的功能紧密相关；- 同时只能有一个
kube-scheduler、kube-controller-manager进程处于工作状态，如果运行多个，则需要通过选举产生一个 leader；
本文档记录部署一个三个节点的高可用 kubernetes master 集群步骤。（后续创建一个 load balancer 来代理访问 kube-apiserver 的请求）
TLS 证书文件
pem和token.csv证书文件我们在TLS证书和秘钥这一步中已经创建过了。我们再检查一下。
<code class="language-bash hljs ">$ ls /etc/kubernetes/ssl admin-key.pem admin.pem ca-key.pem ca.pem kube-proxy-key.pem kube-proxy.pem kubernetes-key.pem kubernetes.pem </code>
www#gaodaima.com来源gaodai^.ma#com搞#代!码网搞代码
下载最新版本的二进制文件
有两种下载方式
方式一
从 github release 页面 下载发布版 tarball，解压后再执行下载脚本
<code class="language-shell hljs ruby"><span class="hljs-variable">$ </span>wget <span class="hljs-symbol">https:</span>/<span class="hljs-regexp">/github.com/kubernetes</span><span class="hljs-regexp">/kubernetes/releases</span><span class="hljs-regexp">/download/v</span>1.<span class="hljs-number">6.0</span>/kubernetes.tar.gz <span class="hljs-variable">$ </span>tar -xzvf kubernetes.tar.gz ... <span class="hljs-variable">$ </span>cd kubernetes <span class="hljs-variable">$ </span>./cluster/get-kube-binaries.sh ... </code>
方式二
从 CHANGELOG页面 下载 client 或 server tarball 文件
server 的 tarball kubernetes-server-linux-amd64.tar.gz 已经包含了 client(kubectl) 二进制文件，所以不用单独下载kubernetes-client-linux-amd64.tar.gz文件；
<code class="language-shell hljs ruby"><span class="hljs-variable">$ </span><span class="hljs-comment"># wget https://dl.k8s.io/v1.6.0/kubernetes-client-linux-amd64.tar.gz</span> <span class="hljs-variable">$ </span>wget <span class="hljs-symbol">https:</span>/<span class="hljs-regexp">/dl.k8s.io/v</span>1.<span class="hljs-number">6.0</span>/kubernetes-server-linux-amd64.tar.gz <span class="hljs-variable">$ </span>tar -xzvf kubernetes-server-linux-amd64.tar.gz ... <span class="hljs-variable">$ </span>cd kubernetes <span class="hljs-variable">$ </span>tar -xzvf kubernetes-src.tar.gz </code>
将二进制文件拷贝到指定路径
<code class="language-bash hljs ">$ cp -r server/bin/{kube-apiserver,kube-controller-manager,kube-scheduler,kubectl,kube-proxy,kubelet} /root/local/bin/
</code>
配置和启动 Kube-Apiserver
创建 kube-apiserver的service配置文件
serivce配置文件/usr/lib/systemd/system/kube-apiserver.service内容：
<code class="language-ini hljs "><span class="hljs-title">[Unit]</span> <span class="hljs-setting">Description=<span class="hljs-value">Kubernetes API Service</span></span> <span class="hljs-setting">Documentation=<span class="hljs-value">https://github.com/GoogleCloudPlatform/kubernetes</span></span> <span class="hljs-setting">After=<span class="hljs-value">network.target</span></span> <span class="hljs-setting">After=<span class="hljs-value">etcd.service</span></span> <span class="hljs-title">[Service]</span> <span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/config</span></span> <span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/apiserver</span></span> <span class="hljs-setting">ExecStart=<span class="hljs-value">/usr/bin/kube-apiserver </span></span> $KUBE_LOGTOSTDERR $KUBE_LOG_LEVEL $KUBE_ETCD_SERVERS $KUBE_API_ADDRESS $KUBE_API_PORT $KUBELET_PORT $KUBE_ALLOW_PRIV $KUBE_SERVICE_ADDRESSES $KUBE_ADMISSION_CONTROL $KUBE_API_ARGS <span class="hljs-setting">Restart=<span class="hljs-value"><span class="hljs-keyword">on</span>-failure</span></span> <span class="hljs-setting">Type=<span class="hljs-value">notify</span></span> <span class="hljs-setting">LimitNOFILE=<span class="hljs-value"><span class="hljs-number">65536</span></span></span> <span class="hljs-title">[Install]</span> <span class="hljs-setting">WantedBy=<span class="hljs-value">multi-user.target</span></span> </code>
/etc/kubernetes/config文件的内容为：
<code class="language-ini hljs ">### # kubernetes system config # # The following values are used to configure various aspects of all # kubernetes services, including # # kube-apiserver.service # kube-controller-manager.service # kube-scheduler.service # kubelet.service # kube-proxy.service # logging to stderr means we get it in the systemd journal <span class="hljs-setting">KUBE_LOGTOSTDERR=<span class="hljs-value"><span class="hljs-string">"--logtostderr=true"</span></span></span> # journal message level, 0 is debug <span class="hljs-setting">KUBE_LOG_LEVEL=<span class="hljs-value"><span class="hljs-string">"--v=0"</span></span></span> # Should this cluster be allowed to run privileged docker containers <span class="hljs-setting">KUBE_ALLOW_PRIV=<span class="hljs-value"><span class="hljs-string">"--allow-privileged=true"</span></span></span> # How the controller-manager, scheduler, and proxy find the apiserver #KUBE_MASTER="--master=http://sz-pg-oam-docker-test-001.tendcloud.com:8080" <span class="hljs-setting">KUBE_MASTER=<span class="hljs-value"><span class="hljs-string">"--master=http://172.20.0.113:8080"</span></span></span> </code>
该配置文件同时被kube-apiserver、kube-controller-manager、kube-scheduler、kubelet、kube-proxy使用。
apiserver配置文件/etc/kubernetes/apiserver内容为：
<code class="language-Ini hljs bash"><span class="hljs-comment">###</span> <span class="hljs-comment">## kubernetes system config</span> <span class="hljs-comment">##</span> <span class="hljs-comment">## The following values are used to configure the kube-apiserver</span> <span class="hljs-comment">##</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## The address on the local server to listen to.</span> <span class="hljs-comment">#KUBE_API_ADDRESS="--insecure-bind-address=sz-pg-oam-docker-test-001.tendcloud.com"</span> KUBE_API_ADDRESS=<span class="hljs-string">"--advertise-address=172.20.0.113 --bind-address=172.20.0.113 --insecure-bind-address=172.20.0.113"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## The port on the local server to listen on.</span> <span class="hljs-comment">#KUBE_API_PORT="--port=8080"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## Port minions listen on</span> <span class="hljs-comment">#KUBELET_PORT="--kubelet-port=10250"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## Comma separated list of nodes in the etcd cluster</span> KUBE_ETCD_SERVERS=<span class="hljs-string">"--etcd-servers=https://172.20.0.113:2379,172.20.0.114:2379,172.20.0.115:2379"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## Address range to use for services</span> KUBE_SERVICE_ADDRESSES=<span class="hljs-string">"--service-cluster-ip-range=10.254.0.0/16"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## default admission control policies</span> KUBE_ADMISSION_CONTROL=<span class="hljs-string">"--admission-control=ServiceAccount,NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota"</span> <span class="hljs-comment">#</span> <span class="hljs-comment">## Add your own!</span> KUBE_API_ARGS=<span class="hljs-string">"--authorization-mode=RBAC --runtime-config=rbac.authorization.k8s.io/v1beta1 --kubelet-https=true --experimental-bootstrap-token-auth --token-auth-file=/etc/kubernetes/token.csv --service-node-port-range=30000-32767 --tls-cert-file=/etc/kubernetes/ssl/kubernetes.pem --tls-private-key-file=/etc/kubernetes/ssl/kubernetes-key.pem --client-ca-file=/etc/kubernetes/ssl/ca.pem --service-account-key-file=/etc/kubernetes/ssl/ca-key.pem --etcd-cafile=/etc/kubernetes/ssl/ca.pem --etcd-certfile=/etc/kubernetes/ssl/kubernetes.pem --etcd-keyfile=/etc/kubernetes/ssl/kubernetes-key.pem --enable-swagger-ui=true --apiserver-count=3 --audit-log-maxage=30 --audit-log-maxbackup=3 --audit-log-maxsize=100 --audit-log-path=/var/lib/audit.log --event-ttl=1h"</span> </code>
--authorization-mode=RBAC指定在安全端口使用 RBAC 授权模式，拒绝未通过授权的请求；- kube-scheduler、kube-controller-manager 一般和 kube-apiserver 部署在同一台机器上，它们使用非安全端口和 kube-apiserver通信;
- kubelet、kube-proxy、kubectl 部署在其它 Node 节点上，如果通过安全端口访问 kube-apiserver，则必须先通过 TLS 证书认证，再通过 RBAC 授权；
- kube-proxy、kubectl 通过在使用的证书里指定相关的 User、Group 来达到通过 RBAC 授权的目的；
- 如果使用了 kubelet TLS Boostrap 机制，则不能再指定
--kubelet-certificate-authority、--kubelet-client-certificate和--kubelet-client-key选项，否则后续 kube-apiserver 校验 kubelet 证书时出现 ”x509: certificate signed by unknown authority“ 错误； --admission-control值必须包含ServiceAccount；--bind-address不能为127.0.0.1；runtime-config配置为rbac.authorization.k8s.io/v1beta1，表示运行时的apiVersion；--service-cluster-ip-range指定 Service Cluster IP 地址段，该地址段不能路由可达；- 缺省情况下 kubernetes 对象保存在 etcd
/registry路径下，可以通过--etcd-prefix参数进行调整；
完整 unit 见 kube-apiserver.service
启动kube-apiserver
<code class="language-bash hljs ">$ systemctl daemon-reload $ systemctl enable kube-apiserver $ systemctl start kube-apiserver $ systemctl status kube-apiserver </code>
配置和启动 Kube-Controller-Manager
创建 kube-controller-manager的serivce配置文件
文件路径/usr/lib/systemd/system/kube-controller-manager.service
<code class="language-ini hljs "><span class="hljs-setting">Description=<span class="hljs-value">Kubernetes Controller Manager</span></span> <span class="hljs-setting">Documentation=<span class="hljs-value">https://github.com/GoogleCloudPlatform/kubernetes</span></span> <span class="hljs-title">[Service]</span> <span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/config</span></span> <span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/controller-manager</span></span> <span class="hljs-setting">ExecStart=<span class="hljs-value">/usr/bin/kube-controller-manager </span></span> $KUBE_LOGTOSTDERR $KUBE_LOG_LEVEL $KUBE_MASTER $KUBE_CONTROLLER_MANAGER_ARGS <span class="hljs-setting">Restart=<span class="hljs-value"><span class="hljs-keyword">on</span>-failure</span></span> <span class="hljs-setting">LimitNOFILE=<span class="hljs-value"><span class="hljs-number">65536</span></span></span> <span class="hljs-title">[Install]</span> <span class="hljs-setting">WantedBy=<span class="hljs-value">multi-user.target</span></span> </code>
配置文件/etc/kubernetes/controller-manager。
<code class="language-ini hljs ">### # The following values are used to configure the kubernetes controller-manager # defaults from config and apiserver should be adequate # Add your own! <span class="hljs-setting">KUBE_CONTROLLER_MANAGER_ARGS=<span class="hljs-value"><span class="hljs-string">"--address=127.0.0.1 --service-cluster-ip-range=10.254.0.0/16 --cluster-name=kubernetes --cluster-signing-cert-file=/etc/kubernetes/ssl/ca.pem --cluster-signing-key-file=/etc/kubernetes/ssl/ca-key.pem --service-account-private-key-file=/etc/kubernetes/ssl/ca-key.pem --root-ca-file=/etc/kubernetes/ssl/ca.pem --leader-elect=true"</span></span></span> </code>
--service-cluster-ip-range参数指定 Cluster 中 Service 的CIDR范围，该网络在各 Node 间必须路由不可达，必须和 kube-apiserver 中的参数一致；--cluster-signing-*指定的证书和私钥文件用来签名为 TLS BootStrap 创建的证书和私钥；--root-ca-file用来对 kube-apiserver 证书进行校验，指定该参数后，才会在Pod 容器的 ServiceAccount 中放置该 CA 证书文件；--address值必须为127.0.0.1，因为当前 kube-apiserver 期望 scheduler 和 controller-manager 在同一台机器，否则：
<code class="language-bash hljs "> $ kubectl get componentstatuses
NAME STATUS MESSAGE ERROR
scheduler Unhealthy Get http://<span class="hljs-number">127.0</span>.<span class="hljs-number">0.1</span>:<span class="hljs-number">10251</span>/healthz: dial tcp <span class="hljs-number">127.0</span>.<span class="hljs-number">0.1</span>:<span class="hljs-number">10251</span>: getsockopt: connection refused
controller-manager Healthy ok
etcd-<span class="hljs-number">2</span> Unhealthy Get http://<span class="hljs-number">172.20</span>.<span class="hljs-number">0.113</span>:<span class="hljs-number">2379</span>/health: malformed HTTP response <span class="hljs-string">"x15x03x01x00x02x02"</span>
etcd-<span class="hljs-number">0</span> Healthy {<span class="hljs-string">"health"</span>: <span class="hljs-string">"true"</span>}
etcd-<span class="hljs-number">1</span> Healthy {<span class="hljs-string">"health"</span>: <span class="hljs-string">"true"</span>}
</code>
参考：https://github.com/kubernetes-incubator/bootkube/issues/64
完整 unit 见 kube-controller-manager.service
启动 Kube-Controller-Manager
<code class="language-bash hljs ">$ systemctl daemon-reload $ systemctl enable kube-controller-manager $ systemctl start kube-controller-manager </code>
配置和启动 Kube-Scheduler
创建 kube-scheduler的serivce配置文件
文件路径/usr/lib/systemd/system/kube-scheduler.serivce。
<code class="language-ini hljs "><span class="hljs-title">[Unit]</span>
<span class="hljs-setting">Description=<span class="hljs-value">Kubernetes Scheduler Plugin</span></span>
<span class="hljs-setting">Documentation=<span class="hljs-value">https://github.com/GoogleCloudPlatform/kubernetes</span></span>
<span class="hljs-title">[Service]</span>
<span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/config</span></span>
<span class="hljs-setting">EnvironmentFile=<span class="hljs-value">-/etc/kubernetes/scheduler</span></span>
<span class="hljs-setting">ExecStart=<span class="hljs-value">/usr/bin/kube-scheduler </span></span>
$KUBE_LOGTOSTDERR
$KUBE_LOG_LEVEL
$KUBE_MASTER
$KUBE_SCHEDULER_ARGS
<span class="hljs-setting">Restart=<span class="hljs-value"><span class="hljs-keyword">on</span>-failure</span></span>
<span class="hljs-setting">LimitNOFILE=<span class="hljs-value"><span class="hljs-number">65536</span></span></span>
<span class="hljs-title">[Install]</span>
<span class="hljs-setting">WantedBy=<span class="hljs-value">multi-user.target</span></span>
</code>
配置文件/etc/kubernetes/scheduler。
<code class="language-Ini hljs bash"><span class="hljs-comment">###</span> <span class="hljs-comment"># kubernetes scheduler config</span> <span class="hljs-comment"># default config should be adequate</span> <span class="hljs-comment"># Add your own!</span> KUBE_SCHEDULER_ARGS=<span class="hljs-string">"--leader-elect=true --address=127.0.0.1"</span> </code>
--address值必须为127.0.0.1，因为当前 kube-apiserver 期望 scheduler 和 controller-manager 在同一台机器；
完整 unit 见 kube-scheduler.service
启动 Kube-Scheduler
<code class="language-bash hljs ">$ systemctl daemon-reload $ systemctl enable kube-scheduler $ systemctl start kube-scheduler </code>
验证 Master 节点功能
<code class="language-bash hljs ">$ kubectl get componentstatuses
NAME STATUS MESSAGE ERROR
scheduler Healthy ok
controller-manager Healthy ok
etcd-<span class="hljs-number">0</span> Healthy {<span class="hljs-string">"health"</span>: <span class="hljs-string">"true"</span>}
etcd-<span class="hljs-number">1</span> Healthy {<span class="hljs-string">"health"</span>: <span class="hljs-string">"true"</span>}
etcd-<span class="hljs-number">2</span> Healthy {<span class="hljs-string">"health"</span>: <span class="hljs-string">"true"</span>}
</code>
后记
当时在配置过程中遇到了问题TLS认证相关的问题，其实就是因为配置apiserver时候etcd的协议写成了http导致的，应该是用https。
Opsnull写的kubernetes高可用master集群部署过程中似乎并没有包括高可用的配置，才云科技的唐继元分享过Kubernetes Master High Availability 高级实践。
究竟如何实现kubernetes master的高可用还需要继续探索。
搞代码网（gaodaima.com）提供的所有资源部分来自互联网，如果有侵犯您的版权或其他权益，请说明详细缘由并提供版权或权益证明然后发送到邮箱[email protected]，我们会在看到邮件的第一时间内为您处理，或直接联系QQ：872152909。本网站采用BY-NC-SA协议进行授权
转载请注明原文链接：Kubernetes高可用Master节点安装
转载请注明原文链接：Kubernetes高可用Master节点安装
