microservice-workshop

This repository hosts the computational environment required to run the following repos:

• https://github.com/ml-starter-packs/flask-backend
• https://github.com/ml-starter-packs/cgi-backend
• https://github.com/ml-starter-packs/flask-frontend

by leveraging mybinder.org, a free service which provides an ephemeral cloud computing environment.

Developer Environments

In your Terminal, you may want to run PS1=\$\ , with echo PS1=\"\$\ \" >> ~/.bashrc for it to persist.

/proxy/.../

To visit services (html pages), change your url to end with {unique-binder-id}/proxy/{PORT}/ (trailing slash included).

/vscode

Code-Server:

[Cmd/Ctrl] + Shift + P -> Add Folder to Workspace -> [Enter], and your home directory will show up on the left.

To open the Terminal, Ctrl + ^.

/lab

Lab:

Click on the badge above and you will be launched into a jupyter session. Your file system is visible on the left-panel, and "apps" are available in the center. Open a Terminal and start checking out the example repos!

/tree

Tree:

Classic Notebook view. Use the button on the top right: Run > Terminal, but note that multiple terminals (which you may want) will require multiple tabs to view.

Running with Docker (locally)

This project was set up to work with both mybinder.org and a local docker build which will launch code-server by default. To use it, run

docker build -t microservices-demo .
docker run --name demo-env -p 5000:5000 microservices-demo

or using docker-compose.yml:

version: "3"

services:
  demo:
    container_name: demo-env
    image: microservices-demo
    build:
      dockerfile: Dockerfile
      context: .
    ports:
      - "5000:5000"
    command: code-server --auth none --bind-addr 0.0.0.0 --port 5000

If you wish to use jupyter lab / notebook, change the command to:

jupyter lab --ip 0.0.0.0 --port 5000 or jupyter notebook --ip 0.0.0.0 --port 5000

and take note of the URL printed in your Terminal to handle authentication with the web application.