diff --git a/README.md b/README.md index e0d60a8..af08b56 100644 --- a/README.md +++ b/README.md @@ -58,11 +58,34 @@ cd kube-deploy/docker-multinode/ 系统后给node打标签,运行 kube-deploy/label.sh 脚本,通过hostname更好识别主机。 -## 运行HADOOP +## 部署私有仓库Harbor + +``` +kube-deploy/easy-rsa.sh + +vi kube-deploy/harbor-make/harbor.cfg +hostname = DOMAIN +ui_url_protocol = https +ssl_cert = crt PATH +ssl_cert_key = key PATH + +# 把yaml里面的配置修改成你本地对应的镜像名 +find kubernetes/ -name "*.rc.yaml" +kubernetes/nginx/nginx.rc.yaml +kubernetes/mysql/mysql.rc.yaml +kubernetes/registry/registry.rc.yaml +kubernetes/ui/ui.rc.yaml +kubernetes/jobservice/jobservice.rc.yaml + +kube-deploy/harbor_k8s +``` + +## 编译部署HADOOP 首选需要制作镜像,有参考网上的Dockerfile的写法。同时结合后面的yaml配置,使用ConfigMap的方式来部署配置和真正的运行脚本: 第一步: 下载JDK、HADOOP + 第二步: 根据镜像的版本编译HADOOP ```