How to Containerize Python Web Applications

Updated on April 22, 2024

Introduction

Containerizing Python web applications encapsulates the application structure and dependencies into a single package to ensure consistent and reproducible deployments across different environments. Deploying containerized Python web applications is a fundamental approach to modern software development that involves scaling and automation when deploying to platforms such as Kubernetes.

You can build Python applications using different development frameworks. Each platform may require a slightly different approach to containerize and ship the application image. Popular Python frameworks include:

Flask: A lightweight, flexible web framework that provides essential tools for building web applications. It's minimalistic and offers direct tools that allow developers to add libraries based on the specific project needs.
Django: A high-level full-stack web framework that provides a complete set of tools and libraries for building robust web applications.
FastAPI: A modern, high-performance web framework for building APIs with Python. It leverages type hints to enable auto-generated interactive API documentation (using Swagger UI) and provides high performance using Python's type system.
Pyramid: A flexible and scalable framework that allows developers to choose the components they need for their applications.
Tornado: An asynchronous networking library and web framework primarily known for its high performance and ability to handle many concurrent connections.

This article explains how to containerize Python web applications for deployment in production environments such as Kubernetes clusters.

Prerequisites

Before you begin:

Deploy a Vultr Kubernetes Engine (VKE) cluster to use as the production deployment platform.
Create a Vultr Container Registry to store the Python application image.
Deploy a One-Click Docker instance to use as the management workstation.
Access the workstation using SSH as a non-root user with sudo privileges.
Install and configure Kubectl to access the cluster.
Install the Helm CLI tool to install additional cluster applications.
console
```
$ sudo snap install helm --classic
```

Create a Python Web Application

To create a consistent Python web application image that you can use with different frameworks such as Flask or Django, structure your code to separate the frontend HTTP server interface logic from the main application logic. Follow the steps below to create a basic Python application using app.py as the main application file while using a server.py file to include the basic HTTP server logic.

Install the Python virtual environment and the PIP package manager.
console
```
$ sudo apt install python3-venv python3-pip -y
```
Create a new project directory to store the application files.
console
```
$ mkdir python-app
```
Switch to the directory.
console
```
$ cd python-app
```
Create a new Python virtual environment to separate the application dependencies from other system packages.
console
```
$ python3 -m venv python-env
```
Activate the virtual environment.
console
```
$ source python-env/bin/activate
```
Create the main application file app.py using a text editor such as Nano.
console
```
(python-env)$ nano app.py
```
Add the following code to the file.
python
```
def hello_world():
 print("Hello, World!")

if __name__ == "__main__":
 hello_world()
```
Save and close the file.

The above code creates a basic Python web application that displays a Hello, World! prompt. Within the application:
- The hello_world function prints the Hello, World! string when called.
- if __name__ == "__main__":: Checks if the script runs as the main program. This ensures that certain code elements run when the script starts with the main thread. If the script is the main program, it calls the hello_world function.
Create a new server.py file to set up a basic HTTP application using the Python http.server and socketserver modules.
console
```
(python-env)$ nano server.py
```

Add the following code to the file.

                            python
                            
                        
from http.server import SimpleHTTPRequestHandler
from socketserver import TCPServer
from threading import Thread

class HelloWorldHandler(SimpleHTTPRequestHandler):
 def do_GET(self):
     # Handle GET requests
     self.send_response(200)
     self.send_header("Content-type", "text/plain")
     self.end_headers()
     self.wfile.write(b"Hello, World!")

def start_server():
 # Set up the HTTP server with custom handler
 server_address = ("", 8000)
 httpd = TCPServer(server_address, HelloWorldHandler)

 print("Server is running at http://localhost:8000")

 try:
     # Serve indefinitely
     httpd.serve_forever()
 except KeyboardInterrupt:
     # Handle keyboard interrupt to gracefully shut down the server
     print("Shutting down the server.")
     httpd.shutdown()

if __name__ == "__main__":
 # Start the server in a separate thread
 server_thread = Thread(target=start_server)
 server_thread.start()

 # Run the main application
 from app import hello_world
 hello_world()

 # Wait for the server thread to finish
 server_thread.join()

Save and close the file.

The above application code responds to incoming GET requests with a Hello, World! prompt. Within the code:

HelloWorldHandler: Defines a custom handler class that inherits the SimpleHTTPRequestHandler HTTP module function to handle GET requests.
start_server: Creates a TCPServer instance running on the host port 8000. In addition, it captures KeyboardInterrupt exceptions Ctrl + C interrupts to gracefully shut down the server.
(if __name__ == "__main__":): Starts the HTTP server in a separate thread using the Thread(target=start_server) function. Then, it imports and calls a hello_world function from the app module. server_thread.join() waits for the server thread to complete before closing the connection.

Start the Python application as a background task in your server session.
console
```
(python-env)$ python3 server.py &
```
Output:
```
Server is running at http://localhost:8000
```
Send a GET request to the host port 8000 to verify access to the application.
console
```
(python-env)$ curl -X GET -H "Content-Type: application/json" http://localhost:8000
```
Output:
```
Hello, World!
```
When the above GET request fails, verify that no conflicting applications run on the defined host port 8000.
View the background application Job ID.
console
```
$ jobs
```
Output:
```
[1] python3 server.py &
```
Stop the application by Job ID. For example 1.
console
```
(python-env)$ kill %1
```
Deactivate your virtual environment.
console
```
(python-env)$ deactivate
```

Containerize the Python Web Application

Create a new Dockerfile Dockerfile to set up your Python application container image.
console
```
$ nano Dockerfile
```
Add the following contents to the file.
Dockerfile
```
FROM python:3.8-slim
WORKDIR /app

COPY . /app

CMD ["python3", "./server.py"]
```
Save and close the file.

The above Dockerfile configuration defines the Python container directory and run-time structure. Within the configuration:
- FROM python:3.8-slim: Uses the official Python 3.8 image as the base image to run the container.
- WORKDIR /app: Sets the working directory to /app inside the container.
- COPY . /app: Copies data from your project directory to the container.
- CMD ["python3", "./server.py"]: Runs the server.py script when the container starts.
Build the Docker image with all directory files.
console
```
$ docker build -t python-app .
```

View the local Docker images to verify that the application is available.

console

$ docker images

Output:

REPOSITORY                         TAG         IMAGE ID       CREATED          SIZE
python-app                               latest             86ff1e8041fe   19 seconds ago   180MB

Push the Python Application Image to the Vultr Container Registry

To store and use the Python application image on a platform such as Kubernetes, push the image to the Vultr Container Registry to ship it to your target environments as described in the steps below.

Log in to your Vultr Container Registry. Replace pythonregistry, vcr-user, vcr-password with your actual registry details.
console
```
$ docker login https://sjc.vultrcr.com/pythonregistry -u vcr-user -p vcr-password
```
Tag the local Docker image with your target Vultr Container Registry repository.
console
```
$ docker tag python-app:latest sjc.vultrcr.com/pythonregistry/python-app:latest
```

View the list of available Docker images and verify that the registry image is available.

console

$ docker images

Output:

sjc.vultrcr.com/pythonregistry/python-app    latest             5594b65d7cd5   3 minutes ago   159MB
python-app                                   latest             01acb92c461c   3 minutes ago   159MB
python                                       3.8-slim           067655fb1c09   2 months ago    128MB

Push your Python application image to the Vultr Container Registry.

                            console
                            
$ docker push sjc.vultrcr.com/pythonregistry/python-app:latest

Output:

The push refers to repository [sjc.vultrcr.com/pythonregistry/python-app]
1.2e1e26d6f: Pushed
1.df21a05: Pushed
b081d565ce24: Pushed
e67a1acfc698: Pushed
1.8551e70: Pushed
da5d55102092: Pushed
fb1bd2fc5282: Pushed
latest: digest: sha256:4a523aab91202e4d79d9456672571ae855a7227590d2dd63de11498bf64f098c size: 1788

Start a new Docker container using your registry container image to test access to the application.
console
```
$ docker run -dit -p 8000:8000 sjc.vultrcr.com/pythonregistry/python-app:latest
```

View the list of running Docker containers to verify that your application container is available.

console

$ docker ps

Output:

CONTAINER ID   IMAGE               COMMAND                 CREATED         STATUS         PORTS                                       NAMES
cfaf3ab6bfac   python-app:latest   "python3 ./server.py"   8 seconds ago   Up 7 seconds   0.0.0.0:8000->8000/tcp, :::8000->8000/tcp   jovial_brattain

Send a new GET request to the host port 8000 to verify access to the application container.
```
$ curl -X GET -H "Content-Type: application/json" http://127.0.0.1:8000
```
Output:
```
Hello, World!
```

Deploy the Python Application in a Kubernetes Cluster

To securely deploy your containerized Python application in a production environment such as a Kubernetes cluster, set up access to your Vultr Container Registry and install the application image as a new Deployment resource. When deployed, your application container listens for incoming connections on port 8000 as defined in the container configuration. Follow the steps below to deploy and use your Python application image in a Vultr Kubernetes Engine (VKE) cluster.

Access your Vultr Container Registry control panel.
Navigate to the Docker/Kubernetes tab, and click Generate Kubernetes YAML to generate a new registry Secret resource configuration.
Create a new Secret resource file secret.yaml.
console
```
$ nano secret.yaml
```

Add your generated Vultr Container Registry YAML configuration to the file similar to the one below:

apiVersion: v1
kind: Secret
metadata:
  name: vultr-cr-credentials
data:
  .dockerconfigjson: example-json
type: kubernetes.io/dockerconfigjson

Save and close the file.

Deploy the Secret to your cluster.
console
```
$ kubectl apply -f secret.yaml
```

View the cluster resources to verify that your registry credentials are ready to use.

                            console
                            
$ kubectl get secret

Output:

NAME                   TYPE                             DATA   AGE
vultr-cr-credentials   kubernetes.io/dockerconfigjson   1      29m

Create a new Deployment resource file deployment.yaml to describe how to run your application in a pod.
console
```
$ nano deployment.yaml
```

Add the following configurations to the file. Replace sjc.vultrcr.com/pythonregistry/python-app:latest with your actual Vultr Container Registry repository.

                            yaml
                            
                        
kind: Deployment
apiVersion: apps/v1
metadata:
 name: python-app
spec:
 replicas: 3
 selector:
   matchLabels:
     app: python-app
 template:
   metadata:
     labels:
       app: python-app
   spec:
     containers:
     - name: python-app
       image: sjc.vultrcr.com/pythonregistry/python-app:latest
       ports:
       - containerPort: 8000
     imagePullSecrets:
     - name: vultr-cr-credentials

Save and close the file.

Apply the deployment to your cluster.
console
```
$ kubectl apply -f deployment.yaml
```

View the cluster deployments to verify that the new resource is available.

                            console
                            
$ kubectl get deployment

Output:

NAME         READY   UP-TO-DATE   AVAILABLE   AGE
python-app   3/3     3            3           21s

View the cluster pods to verify all pods associated with your deployment.

                            console
                            
$ kubectl get pods

Output:

NAME                          READY   STATUS    RESTARTS   AGE
python-app-6cb76cfc8d-5wjbf   1/1     Running   0          35s
python-app-6cb76cfc8d-tphrx   1/1     Running   0          35s
python-app-6cb76cfc8d-txk5p   1/1     Running   0          35s

Securely Expose the Python Application for External Access

To Expose your Python application for external access using your Vultr Kubernetes Engine (VKE) cluster, install an Ingress controller to map your application service to a domain name. The controller securely routes all application requests from your domain name to the backend Python application service. Follow the steps below to expose the application in your VKE cluster.

Add the NGINX ingress repository to your Helm sources.

                            console
                            
$ helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx

Update the Helm repository index.
console
```
$ helm repo update
```

Install the NGINX Ingress Controller to your cluster using Helm.

                            console
                            
$ helm install nginx-ingress ingress-nginx/ingress-nginx --set controller.publishService.enabled=true

Wait for at least 3 minutes for the Nginx Ingress Controller to install all necessary CRDs. Then, view the controller services to verify the assigned load balancer IP address.

                            console
                            
$ kubectl get service --namespace default nginx-ingress-ingress-nginx-controller

Output:

NAME                                     TYPE           CLUSTER-IP     EXTERNAL-IP     PORT(S)                      AGE
nginx-ingress-ingress-nginx-controller   LoadBalancer   10.98.94.149   192.0.2.100   80:32671/TCP,443:32325/TCP   2m58s

Keep note of the Ingress controller EXTERNAL-IP value to externally access your cluster services.

Set Up a new domain A record pointing to the Ingress controller public IP address. For example, app.example.com.
Create a new Python Service resource file service.yaml
console
```
$ nano service.yaml
```

Add the following contents to the file.

                            yaml
                            
                        
apiVersion: v1
kind: Service
metadata:
 name: python-app
spec:
 selector:
   app: python-app
 ports:
   - protocol: TCP
     port: 80
     targetPort: 8000

Save and close the file.

The above Service configuration forwards all network requests on the HTTP port 80 to your Python application port 8000.

Apply the Service to your cluster.
console
```
$ kubectl apply -f service.yaml
```
Create a new Ingress resource file ingress.yaml to expose the Python application service.
console
```
$ nano ingress.yaml
```

Add the following contents to the file. Replace app.example.com with your actual domain,

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: nginx-ingress
  namespace: default
  annotations:
    kubernetes.io/ingress.class: nginx
spec:
  rules:
  - host: app.example.com
    http:
      paths:
      - backend:
          service:
            name: python-app
            port:
              number: 80
        path: /
        pathType: Prefix

Save and close the file.

The above Ingress configuration forwards all root requests from app.example.com to your Python application service on port 80.

Apply the Ingress resource to your cluster.
console
```
$ kubectl apply -f ingress.yaml
```

View the cluster Ingress objects and verify that the new resource is available.

                            console
                            
$ kubectl get ingress

Output:

NAME            CLASS   HOSTS              ADDRESS         PORTS   AGE
nginx-ingress   nginx   app.example.com   192.0.2.100   80      32s

Access your domain using a web browser such as Firefox to test access to the application.

Conclusion

You have containerized a Python application and stored it in a Vultr Container Registry repository for deployment in a VKE cluster. Shipping Python web applications as container images enables the efficient management and upgrade of application components for deployment in multiple environments.

How to Containerize Python Web Applications

Introduction

Prerequisites

Create a Python Web Application

Containerize the Python Web Application

Push the Python Application Image to the Vultr Container Registry

Deploy the Python Application in a Kubernetes Cluster

Securely Expose the Python Application for External Access

Conclusion

Products

Features

Marketplace

Resources

Company