Building a Flask(Python) CRUD API with Cloud Firestore(Firebase) and Deploying on Cloud Run.

In this tutorial, you will build a CRUD (Create, Read, Update, Delete) API to manage Todo Lists using Flask (a microframework for Python), Cloud Firestore (a flexible, scalable database for mobile, web, and server development from Firebase) and deploy the API to Cloud Run (a serverless environment to run containers on Google Cloud Platform).

Cloud Firestore stores data as collections of documents, it also features richer, faster queries and scales further than the Realtime Database.
You would be able to manage Todo List documents of varying fields via the API.


Before you begin

  • Create a new Firebase project, or use an existing one.
    • Click on Database and Create database in the Cloud Firestore section.
    • Set your Security Rules and Location
    • You should have an initial screen similar to: Screenshot from 2019-08-28 15-33-32.png
  • Download your Firebase Service Account Key

    • Click on the Settings Icon at the top of your dashboard
    • Click on the Service Account Tab
    • Select Python option for Admin SDK configuration snippet and Click on Generate new private key and save it as key.json Screenshot from 2019-08-28 21-01-04.png
  • Create a new Google Cloud Platform (GCP) project, or use an existing one (you would need this to deploy to Cloud Run)

    • Install Cloud SDK or use Cloud Shell available on the Googl
    • (Optional) To set up Continuous Deployment check out this
    • Ensure you can run gcloud -h on your shell. Screenshot from 2019-08-29 00-01-27.png

Source Code

coding.gif Time to write some codes.


# Required Imports
import os
from flask import Flask, request, jsonify
from firebase_admin import credentials, firestore, initialize_app

# Initialize Flask App
app = Flask(__name__)

# Initialize Firestore DB
cred = credentials.Certificate('key.json')
default_app = initialize_app(cred)
db = firestore.client()
todo_ref = db.collection('todos')

@app.route('/add', methods=['POST'])
def create():
        create() : Add document to Firestore collection with request body
        Ensure you pass a custom ID as part of json body in post request
        e.g. json={'id': '1', 'title': 'Write a blog post'}
        id = request.json['id']
        return jsonify({"success": True}), 200
    except Exception as e:
        return f"An Error Occured: {e}"

@app.route('/list', methods=['GET'])
def read():
        read() : Fetches documents from Firestore collection as JSON
        todo : Return document that matches query ID
        all_todos : Return all documents

        # Check if ID was passed to URL query
        todo_id = request.args.get('id')    
        if todo_id:
            todo = todo_ref.document(todo_id).get()
            return jsonify(todo.to_dict()), 200
            all_todos = [doc.to_dict() for doc in]
            return jsonify(all_todos), 200
    except Exception as e:
        return f"An Error Occured: {e}"

@app.route('/update', methods=['POST', 'PUT'])
def update():
        update() : Update document in Firestore collection with request body
        Ensure you pass a custom ID as part of json body in post request
        e.g. json={'id': '1', 'title': 'Write a blog post today'}
        id = request.json['id']
        return jsonify({"success": True}), 200
    except Exception as e:
        return f"An Error Occured: {e}"

@app.route('/delete', methods=['GET', 'DELETE'])
def delete():
        delete() : Delete a document from Firestore collection

        # Check for ID in URL query
        todo_id = request.args.get('id')
        return jsonify({"success": True}), 200
    except Exception as e:
        return f"An Error Occured: {e}"

port = int(os.environ.get('PORT', 8080))
if __name__ == '__main__':, host='', port=port)

There are individual methods and routes for each action the API would perform, you can improve upon the code snippet and add more functions to meet your needs. For each CRUD (Create, Read, Update, Delete) action, we define the route and its corresponding HTTP Method. We also try to perform that action and return a response with the 200 status code, if there's a problem somewhere, we return the exception's error.

Deploy to Cloud Run

We would have to write a Dockerfile for our API so as to be able to build it as a container and have it run on Cloud Run.

# Dockerfile
FROM python:3.7-stretch
RUN apt-get update -y
RUN apt-get install -y python-pip python-dev build-essential
COPY . /app
RUN pip install -r requirements.txt
ENTRYPOINT ["python"]
CMD [""]

Ensure that you have a requirements.txt file with the following contents.

# requirements.txt

Finally, our cloudbuild.yaml file which you will use to trigger builds needs to be created as follows.

# cloudbuild.yaml
  # build & push the container image
- name: ""
  args: ["--cache=true", "--cache-ttl=48h", "$PROJECT_ID/todo:latest"]
  # Deploy container image to Cloud Run
- name: ""
  args: ['beta', 'run', 'deploy', 'todo', '--image', '$PROJECT_ID/todo:latest', '--region', 'us-central1', '--allow-unauthenticated', '--platform', 'managed']

In addition, you can also import your key.json Firebase service account file into the same directory as it is not advisable to have it pushed to your source repository. A fast approach to this is to add an additional Cloud Build step that downloads the service account from a private location such as a Google Cloud Storage Bucket.

Execute Build & Deploy Steps on Cloud Shell

Run the following command to build your Docker container and push to Container Registry as specified in the cloudbuild.yaml file. This also performs an extra step of deploying to Cloud Run.

gcloud builds submit --config cloudbuild.yaml .

Deploy using Cloud Run Button

Recently, Google Cloud announced Cloud Run Button, an image and link you can add to the README of your source code repositories to allow others to deploy your application to Google Cloud Platform using Cloud Run. The steps to add the Cloud Run Button to your repository are:

  • Copy & paste this markdown into your file
    [![Run on Google Cloud](](
  • Replace YOUR_HTTP_GIT_URL with your HTTP git URL, like:
  • Ensure that your repository has a Dockerfile

Screenshot from 2019-08-29 07-12-49.png

Cleaning up

To prevent unnecessary charges, clean up the resources created for this tutorial, delete the resources or projects used (if you created a new project).

Useful Links

No Comments Yet