Skip to content
Unikraft Cloud

Run a Python app

In this guide we create and deploy a simple Python-based HTTP web server. To run this example, follow these steps:

  1. Install the kraft CLI tool and a container runtime engine, e.g. Docker.

  2. Clone the examples repository and cd into the examples/http-python3.12/ directory:

Terminal window
git clone https://github.com/kraftcloud/examples
cd examples/http-python3.12/

Make sure to log into Unikraft Cloud by setting your token and a metro close to you. We use fra0 (Frankfurt, 🇩🇪) in this guide:

Terminal window
# Set Unikraft Cloud access token
export UKC_TOKEN=token
# Set metro to Frankfurt, DE
export UKC_METRO=fra0

When done, invoke the following command to deploy this application on Unikraft Cloud:

Terminal window
kraft cloud deploy -p 443:8080 -M 512 .

The output shows the instance URL and other details:

Terminal window
[●] Deployed successfully!
 │
 ├────────── name: http-python312-ma2i9
 ├────────── uuid: e7389eee-9808-4152-b2ec-1f3c0541fd05
 ├───────── state: running
 ├─────────── url: https://young-night-5fpf0jj8.fra0.kraft.host
 ├───────── image: http-python312@sha256:278cb8b14f9faf9c2702dddd8bfb6124912d82c11b4a2c6590b6e32fc4049472
 ├───── boot time: 15.09 ms
 ├──────── memory: 512 MiB
 ├─────── service: young-night-5fpf0jj8
 ├── private fqdn: http-python312-ma2i9.internal
 ├──── private ip: 172.16.3.3
 └────────── args: /usr/bin/python /src/server.py

In this case, the instance name is http-python312-ma2i9 and the URL is https://young-night-5fpf0jj8.fra0.kraft.host. They are different for each run.

Use curl to query the Unikraft Cloud instance of the Python-based HTTP web server:

Terminal window
curl https://young-night-5fpf0jj8.fra0.kraft.host
Hello, World!

At any point in time, you can list information about the instance:

Terminal window
kraft cloud instance list
NAME                  FQDN                                  STATE    CREATED AT    IMAGE                                        MEMORY   ARGS                            BOOT TIME
http-python312-ma2i9  young-night-5fpf0jj8.fra0.kraft.host  running  1 minute ago  http-python312@sha256:278cb8b14f9faf9c27...  512 MiB  /usr/bin/python /src/server.py  15094us

When done, you can remove the instance:

Terminal window
kraft cloud instance remove http-python312-ma2i9

Customize your Application

To customize the application, update the files in the repository, listed below:

  • server.py: the actual Python HTTP server
  • Kraftfile: the Unikraft Cloud specification
  • Dockerfile: the Docker-specified application filesystem
import argparse
from http.server import HTTPServer, BaseHTTPRequestHandler


class MyServer(BaseHTTPRequestHandler):
  def do_GET(self):
    self.send_response(200)
    self.send_header("Content-type", "text/html")
    self.end_headers()
    self.wfile.write(bytes("Hello, World!\n", "utf-8"))


def main(args):
  server = HTTPServer((args.host, args.port), MyServer)


  print("starting server at %s:%s" % (args.host, args.port))


  try:
    server.serve_forever()


  except KeyboardInterrupt:
    pass


  print("server stopped")


def parse_args():
  parser = argparse.ArgumentParser()
  parser.add_argument("--host", type=str, default="0.0.0.0")
  parser.add_argument("--port", type=int, default=8080)
  return parser.parse_args()


if __name__ == "__main__":
  main(parse_args())

Lines in the Kraftfile have the following roles:

  • spec: v0.6: The current Kraftfile specification version is 0.6.

  • runtime: python:3.12: The Unikraft runtime kernel to use is Python 3.12.

  • rootfs: ./Dockerfile: Build the application root filesystem using the Dockerfile.

  • cmd: ["/usr/bin/python3", "/src/server.py"]: Use /usr/bin/python3 /src/server.py as the starting command of the instance.

Lines in the Dockerfile have the following roles:

  • FROM scratch: Build the filesystem from the scratch container image, to create a base image.

  • COPY ./server.py /src/server.py: Copy the server implementation file (server.py) in the Docker filesystem (in /src/server.py).

The following options are available for customizing the application:

  • If only updating the implementation in the server.py source file, no other change is required.

  • If new files are added, these have to be copied in the application filesystem, using the COPY command in the Dockerfile.

  • If a new Python source files is added, update the cmd line in the Kraftfile and replace server.py to run that file when creating the instance.

  • More extensive changes may require expanding the Dockerfile with additional Dockerfile commands. This includes the use of Python frameworks and the use of pip, as shown in the next section.

Using pip

pip is a package manager for Python. It is used to install dependencies for Python applications. pip uses the requirements.txt file to list required dependencies (with versions).

The http-python3.12-flask3.0 guide details the use of pip to deploy an application using the Flask framework on Unikraft Cloud.

Run the command below to deploy the application on Unikraft Cloud:

Terminal window
kraft cloud deploy -p 443:8080 -M 512 .

Differences from the http-python3.12 app are also the steps required to create an pip-based app:

  1. Add the requirements.txt file used by pip.

  2. Add framework-specific source files. In our case, this means the server.py file.

  3. Update the Dockerfile to:

    1. COPY the local files.

    2. RUN the pip3 install command to install dependencies.

    3. COPY of the resulting and required files (/usr/local/lib/pyhon3.12 and server.py) in the application filesystem, using the scratch container.

The files are listed below:

from flask import Flask
app = Flask(__name__)


@app.route('/')
def hello():
    return "Hello, World!\n"


if __name__ == '__main__':
    app.run(host='0.0.0.0', port=8080)

The requirements.txt file lists the flask dependency.

The Kraftfile is the same one used for http-python3.12.

For Dockerfile newly added lines have the following roles:

  • FROM python:3.12-bookworm AS base: Use the base image of the python:3.12-bookworm container. This provides the pip3 binary and other Python-related components. Name the current image base.

  • WORKDIR /app: Use /app as working directory. All other commands in the Dockerfile run inside this directory.

  • COPY requirements.txt /app: Copy the package configuration file to the Docker filesystem.

  • RUN pip3 install ...: Install pip components listed in requirements.txt.

  • COPY --from=base ...: Copy generated Python files in the new base image in the scratch-based image.

Similar actions are required for other pip3-based applications. See also the http-python3.12-django5.0 example.

Learn More

Use the --help option for detailed information on using Unikraft Cloud:

Terminal window
kraft cloud --help

Or visit the CLI Reference.