rke搭建k8s集群与rancher高可用部署
作者：互联网
rancher集群部署
在线部署
- 开启rancher服务，UI界面添加集群，集群扩展节点。
docker run -d --privileged --restart=unless-stopped \
-p 80:80 -p 443:443 \
--privileged \
rancher/rancher:latest
- 注意事项
* arm64网络仅支持fannel
* 所有节点root权限
* 节点角色全部选择（crontolplane、etcd、worker）
* etcd集群要求节点数大于三个，因此集群节点不够三个时，集群不能移除节点
* 安装出错需要全部清理重来
离线部署
rke部署k8s集群
创建私有创库
### 加载仓库镜像
docker load registry.tar
### 设置仓库用户密码，用户名admin，密码sysadmin
mkdir -p /opt/registry/auth
docker run --entrypoint htpasswd registry:2.7.0 -Bbn admin sysadmin > /opt/registry/auth/htpasswd
### 开启镜像仓库
docker run -d -p 5000:5000 --restart=always --name my_registry -v /opt/registry/auth:/auth -e "REGISTRY_AUTH=htpasswd" -e "REGISTRY_AUTH_HTPASSWD_REALM=Registry Realm" -e REGISTRY_AUTH_HTPASSWD_PATH=/auth/htpasswd registry:2.7.0
### 验证登录
docker login 127.0.0.1
将所需镜像推送到私有仓库
拉取镜像脚本地址
https://github.com/rancher/rancher/releases/tag/v2.5.5
./rancher-load-images.sh --image-list ./rancher-images.txt --registry 127.0.0.1:5000
安装rke
- 配置免密
ssh-keygen
- rke安装节点需要对集群所有节点免密
ssh-copy-id -i ~/.ssh/id_rsa.pub root@x.x.x.x
- 添加rke运行脚本
cp rke_linux-arm64 /usr/local/bin/rke
- 配置rke集群文件
vi rke-cluster.yml
network:
plugin: flannel
nodes:
- address: 173.21.31.41
user: root ### centos需要docker用户组的非root用户，
role: ["controlplane", "etcd", "worker"]
ssh_key_path: /root/.ssh/id_rsa
- address: 173.21.31.42
user: root
role: ["controlplane", "etcd", "worker"]
ssh_key_path: /root/.ssh/id_rsa
- address: 173.21.31.43
user: root
role: ["controlplane", "etcd", "worker"]
ssh_key_path: /root/.ssh/id_rsa
private_registries: ## 联网环境可以不指定私有仓库
- url: 127.0.0.1:5000
user: admin
password: sysadmin
is_default: true
创建rke集群
rke up --config rke-cluster.yml
- 安装成功会生成kube_config_rke-cluster.yml，rke-cluster.rkestate两个文件
- 同步集群信息
mkdir -p /root/.kube
- 分别在集群所有节点替换/root/.kube/config文件内容与kube_config_rke-cluster.yml内容一致
cp kube_config_rke-cluster.yml /root/.kube/config
- 查看集群信息
kubectl get nodes
NAME STATUS ROLES AGE VERSION
173.21.31.41 Ready controlplane,etcd,worker 27h v1.19.7
173.21.31.42 Ready controlplane,etcd,worker 27h v1.19.7
173.21.31.43 Ready controlplane,etcd,worker 27h v1.19.7
kubectl get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
cattle-system rancher-85c9cf85ff-bm5kl 1/1 Running 80 21h
cattle-system rancher-85c9cf85ff-fs5lc 1/1 Running 4 21h
cattle-system rancher-85c9cf85ff-v5g77 1/1 Running 70 21h
cattle-system rancher-webhook-b5b7b76c4-vx6m8 1/1 Running 1 21h
cert-manager cert-manager-5fdd5877dc-tlddh 1/1 Running 6 23h
cert-manager cert-manager-cainjector-54cbc96f6b-9t5dd 1/1 Running 6 23h
cert-manager cert-manager-webhook-75c9994b75-mlwkb 1/1 Running 1 23h
fleet-system fleet-agent-76fc77855-d78c4 1/1 Running 6 22h
fleet-system fleet-controller-ccc95b8cd-pzpn2 1/1 Running 78 22h
fleet-system gitjob-5997858b9c-bwch9 1/1 Running 7 22h
ingress-nginx default-http-backend-65dd5949d9-9stvt 1/1 Running 2 27h
ingress-nginx nginx-ingress-controller-bng8g 0/1 CrashLoopBackOff 91 27h
ingress-nginx nginx-ingress-controller-pxbhn 1/1 Running 0 27h
ingress-nginx nginx-ingress-controller-qn7w2 1/1 Running 1 27h
kube-system coredns-6f85d5fb88-bls28 1/1 Running 1 27h
kube-system coredns-6f85d5fb88-h4kst 1/1 Running 0 26h
kube-system coredns-6f85d5fb88-zt25r 1/1 Running 1 26h
kube-system coredns-autoscaler-79599b9dc6-xkfsq 1/1 Running 0 27h
kube-system kube-flannel-57k2w 2/2 Running 2 27h
kube-system kube-flannel-mq7m8 2/2 Running 0 27h
kube-system kube-flannel-sljf8 2/2 Running 2 27h
kube-system metrics-server-8449844bf-d5zrz 1/1 Running 1 27h
kube-system rke-coredns-addon-deploy-job-xh9dn 0/1 Completed 0 27h
kube-system rke-ingress-controller-deploy-job-dfgkb 0/1 Completed 0 27h
kube-system rke-metrics-addon-deploy-job-gdhxq 0/1 Completed 0 27h
kube-system rke-network-plugin-deploy-job-n2r6l 0/1 Completed 0 27h
kube-system tiller-deploy-6b6f975bcb-x4f8f 1/1 Running 1 23h
rancher-operator-system rancher-operator-f54cf4887-r68cj 1/1 Running 6 23h
rancher高可用
安装cert_manager
- rancher的自认证模式
# 安装 CustomResourceDefinition 资源
kubectl apply --validate=false -f https://github.com/jetstack/cert-manager/releases/download/v0.15.0/cert-manager.crds.yaml
# 为 cert-manager 创建命名空间
kubectl create namespace cert-manager
# 添加 Jetstack Helm 仓库
helm repo add jetstack https://charts.jetstack.io
# 更新本地 Helm chart 仓库缓存
helm repo update
# 安装 cert-manager Helm chart
helm install \
cert-manager jetstack/cert-manager \
--namespace cert-manager \
--version v0.15.0
- 查看安装情况
kubectl get pods --namespace cert-manager
NAME READY STATUS RESTARTS AGE
cert-manager-5fdd5877dc-tlddh 1/1 Running 6 28h
cert-manager-cainjector-54cbc96f6b-9t5dd 1/1 Running 6 28h
cert-manager-webhook-75c9994b75-mlwkb 1/1 Running 1 28h
安装rancher
- rancher-stable添加到本地
helm repo add rancher-stable https://releases.rancher.com/server-charts/stable
- rancher-stable下载到本地
helm fetch rancher-stable/rancher
- 离线启动
helm template rancher ./rancher-2.3.5.tgz \
--namespace cattle-system --output-dir . \ --set hostname=qx.rancher.com
- 联网启动
helm install rancher rancher-stable/rancher \
--namespace cattle-system \
--set hostname=qx.rancher.com
- 查看启动情况：
kubectl -n cattle-system rollout status deploy/rancher
UI访问
https://qx.rancher.com
用户名：admin，密码：sysadmin
标签：manager,kube,system,rke,Running,rancher,k8s 来源： https://blog.csdn.net/qq_43027317/article/details/113770598