https://github.com/salfaris/streamlit-gcp-minimal

The most minimal MVP for a streamlit web app containerized with Docker and deployed to Google Cloud Run
https://github.com/salfaris/streamlit-gcp-minimal

Last synced: about 9 hours ago
JSON representation

The most minimal MVP for a streamlit web app containerized with Docker and deployed to Google Cloud Run

• Host: GitHub
• URL: https://github.com/salfaris/streamlit-gcp-minimal
• Owner: salfaris
• Created: 2024-07-23T08:36:21.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2024-07-23T10:35:43.000Z (4 months ago)
• Last Synced: 2024-07-24T11:24:26.612Z (4 months ago)
• Language: Shell
• Homepage:
• Size: 169 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Minimal Streamlit app on Google Cloud Run

The most minimal example of a Streamlit app, containerized using Docker and deployed to GCP Cloud Run.

Date written: 23 July 2024

## Step-by-step reproduction

If you want to understand how to build this incrementally, here are the steps.

Note that these steps are ordered and should be done sequentially.

### 0️⃣ Pre-requisites

- [ ] Clone the repo.

- [ ] Install Docker.

- [ ] Create a python environment (e.g. use Conda).

- [ ] Run `pip install -r app/requirements.txt`.

  - Minimally, you should need to install `streamlit`.

### 1️⃣ Creating the Streamlit app

As a first time Docker user, you want to create a Streamlit app and test that it works locally. This way you can be sure any errors are purely due to Python or Streamlit and not Docker.

- [ ] Create a new Streamlit app. Here we built it in `app/app.py`.

### 2️⃣ Dockerize the app

Now we begin the fun part.

- [ ] Create a `Dockerfile`

- [ ] Dockerize

  - [ ] Build the image: `docker build -t  .`

  - [ ] Run the container from image: `docker run `

- [ ] Optionally, create a shell script that combines the dockerize step above. This is `scripts/build_and_run_locally.sh`

### 3️⃣ Pushing image to GCP Artifact Registry

- [ ] Create a GCP project

   - If you already have a GCP project, you might need to switch your project on the terminal. Run `gloud config set project `.

- [ ] Activate the Artifact Registry API

   - Search the Artifact Registry on your GCP Cloud Console and click activate.

 - [ ] Dockerize

   - [ ] Identify the registry host: `eu.gcr.io`

   - [ ] Build the image: `docker build -t / . --platform linux/amd64`

- [ ] Push the image to GCP Artifact Registry: `docker push >`

- [ ] Optionally, create a shell script that combines the two steps above. This is `scripts/build_and_push_to_gcpar.sh`

### 4️⃣ Creating a GCP Cloud Run service

- [ ] Create a GCP Cloud Run service

  - This should automatically enable the Cloud Run service.

  - Creating a service should bring you to a page with this heading: ![create-service](./docs/cloud-run-create-service.png)

- [ ] Specify the `Container image URL` to be the image that was pushed in the previous step.

  - ![choose-container-image](./docs/cloud-run-choose-container-image.png)  

- [ ] Choose your settings and click `CREATE`