AutoML

We're designing the AutoML platform that would make it easy for non-technical people to build custom AI solutions and AI-powered process automation.

Easy-to-use ML platforms leverage the time/value/knowledge trade-off in a genuinely attractive way and allow users with no AI coding skills to optimize day-to-day operations and to solve business issues.

API

API Link: Todo

How Minikube deploys AutoML

AutoML的部署脚本维护在./deploy目录下. 按照以下指南在你本地Minikube上部署AutoML平台.

Prerequisites

开始之前,请确保以下组件安装在你的机器上:

Docker: (参考:https://docs.docker.com/desktop/install/mac-install/）
Minikube: (参考:https://minikube.sigs.k8s.io/docs/start/)
Helm:(参考:https://helm.sh/zh/docs/)

Mac OS:

Minikube在macOS上首选驱动为Docker。

启动Docker
使用 Docker驱动启动集群：minikube start --driver=docker
配置Docker为默认驱动程序：minikube config set driver docker
执行kubectl get node 验证启动是否成功。

Linux: Todo

Windows:

Minikube在Windows上首选驱动为Hyper-V+Docker。我们使用Docker来作为驱动程序。

启动Docker
使用 Docker 驱动启动集群：minikube start --driver=docker
配置 Docker 为默认驱动程序：minikube config set driver docker
执行kubectl get node 验证启动是否成功

加载本地镜像到 minikube: 构建镜像:

docker build -t automl/automl:v1beta1 java/automl

加载本地镜像：

minikube image load automl/automl:v1beta1 # minikube image load <image>, your local image

Start Deploy

配置automl仓库:

helm repo add automl https://ml-for-good.github.io/automl/

搜索automl仓库下的chart:

helm search repo automl

安装automl的helm包：

helm install automl automl/deploy

查看pod,svc,deployment信息:

kubectl get pod,svc,deployment,pv,pvc -n mysql

测试Mysql连接

获取 minikube的 IP 和service的NodePort

minikube service mysql -n mysql --url

使用本地Navicat客户端连接数据库,连接信息:

Host:IP

port:NodePort

用户名:root

密码:root

进入之后, 可以看到库automl和automl-test和相关表已经创建了.

Test OpenAPI mock server

Run service tunnel

minikube service openapi-mock

Try in your browser

Open in your browser (ensure there is no proxy set)

http://127.0.0.1:TUNNEL_PORT/v1beta1/namespaces/1/datasets

Test AUTOML java web demo server

Load local image to minikube

minikube image load automl/automl:v1 # minikube image load <image>, your local image

Run service tunnel

minikube service automl-web

Try in your browser

Open in your browser (ensure there is no proxy set)

http://127.0.0.1:TUNNEL_PORT/swagger-ui/index.html

Get pod, deployment status, name

kubectl get pod,deployment

Get logs

kubectl logs <resource name>

Stop automl-web pod & deployment

kubectl delete deploy automl-web

switch java application link mysql prod.

默认java/automl/目录下的Dockerfile中的 -Dspring.profiles.active=dev 环境是dev环境, 采用automl库

在本地自己测试部署的时候, 可以修改这个值为test这样, automl-web 服务采用automl-test库.

test link

部署完automl-web之后, 本地执行以下命令.

curl -X POST "http://127.0.0.1:TUNNEL_PORT/v1beta1/test/insert" -H "accept: /" -H "Content-Type: application/json" -d "{ "id": 160676, "name": "sss"}"

我们可以看到已经有数据了. (我当时是以test环境启动的)

Todo

jia9iZhang/automl