Talk "Comment la GCP m'a permis de sauver ma Prod chez Mamie"

How to run it Locally

Here's a few steps to build this app and make it run locally. You'll need a running Docker instance and either NodeJS or Python3.

All the following commands are supposed to be run from this directory.

Step 1 : Build the Database and the server image

Note : Note that the datas will not persisted since its only expected to be run locally for dev or tests purpose.

$ cd datas
$ docker build -t gcpmamie-db .
$ cd ../back
$ docker build -t gcpmamie-server .

Step 2 : Run the backend and the database

$ docker-compose up

The backend will listen on `localhost:8081ù

Step 3 : Run the client

If you have NodeJS, you can use thoses commands

$ cd front/
$ npm install http-server
$ http-server -p 8080 

If you have Python3, you can use thoses commands

$ cd front/
$ python3 -m http.server 8080

How to run it on GCP

Prepare your GCP project

Once you've created your GCP project, add the role roles/cloudsql.admin to the Deployement Manager's service account (<projectnumber>@cloudservices.gserviceaccount.com). You also needs to create a Cloud Storage bucket and add your SQL dump in it.

Create the infrastructure

In the cloud shell console (or if you have a gcloud command installed locally), run :

$ cd gcp/
$ make setup

or you can create the infrastructure, resource by resource using :

cd gcp/
$ make install-db

Clean up

If you want to clean up your project, you can either remove everything with :

$ cd gcp/
$ make teardown

or cleaning up resource by resource :

$ cd gcp/
$ make destroy-db

TodoList

Mandatory

• (Front) Generate the products in javascript loop instead of hardcoded
• (Front) Create the Error Page
• (Front) Create a constants.js file which will contains local & GCP url to the backend.
• (Front) Call the backend API to get all the products
• (Front) Call the backend API to add a product to cart
• (Back) Create the Springboot backend Docker image
• (Back) Add an endpoint to return the list of all the products' summaries
• (Back) Add an endpoint to return the json description of a specific product
• (Back) Add an endpoint to add a product in the cart
• (Back) Add an endpoint to get the cart count
• (Data) Create the Database structure in a PostgreSQL instance
• (Data) Fill the database with dummy products
• (Batch) Create a simple dummy/mock batch application which will display random orders informations at each Pub/Sub message
• Create a DM script to organize the setup and teardown of the platform
• Create a makefile to easily setup/teardown and startup/shutdown the resources
• (Front) Deploy on GCS
• (Front) Store the images in GCP CDN
• (Back) Add a cloud build task to build the backend image and push it on GCR
• (Back) Deploy on GKE
• (Back) Add the GCP Endpoints
• (Data) Apply the database structure & init to a GCP CloudSQL
• (Batch) Create the Pub/Sub topic
• (Batch) Deploy 2 instances of the VM batch using an instance template.
• (Batch) Prepare the way to easily replace the 2 hard provisioned VMs to an instance group of 10 VMs (* [ ] (Batch) Prepare the script which will overflow the pubsub topic and the way to easily upgrade the instance group temporarly if previous step failed.)

If there is more time

• (Front) Add the GCP LB in front of the GCS
• (Front) Finish up the main-menu component (with an apparition only on hover) and the menu items in it (Log In, About...)
• (Front) Create the my-account component to display info of a dummy user.
• (Front) Create a search component which do a simple search on the products.
• (Front) Implements the Reset Filter button & add the connection to the backend for the Apply Filter one.
• (Front) Display the description instead of the image when the product is hovered
• (Front) Implement the behavior of the - and + button on the Product page
• (Front) Create the cart popup when hovering the icon which will displays the cart's content, the quantity, the qty modifier, and the cart total price.
• (Front) Create the Checkout Page
• (GCP) Create the dump bucket directly from db-import.py and upload the dump in it.
• (GCP) Script the backend creation in place of a Markdown procedure