Deploying FastKafka using Docker

Building a Docker Image

To build a Docker image for a FastKafka project, we need the following items:

1. A library that is built using FastKafka.
2. A file in which the requirements are specified. This could be a requirements.txt file, a setup.py file, or even a wheel file.
3. A Dockerfile to build an image that will include the two files mentioned above.

Creating FastKafka Code

Let’s create a FastKafka-based application and write it to the application.py file based on the tutorial.

# content of the "application.py" file

from contextlib import asynccontextmanager

from sklearn.datasets import load_iris
from sklearn.linear_model import LogisticRegression

from fastkafka import FastKafka

ml_models = {}


@asynccontextmanager
async def lifespan(app: FastKafka):
    # Load the ML model
    X, y = load_iris(return_X_y=True)
    ml_models["iris_predictor"] = LogisticRegression(random_state=0, max_iter=500).fit(
        X, y
    )
    yield
    # Clean up the ML models and release the resources
    ml_models.clear()


from pydantic import BaseModel, NonNegativeFloat, Field

class IrisInputData(BaseModel):
    sepal_length: NonNegativeFloat = Field(
        ..., example=0.5, description="Sepal length in cm"
    )
    sepal_width: NonNegativeFloat = Field(
        ..., example=0.5, description="Sepal width in cm"
    )
    petal_length: NonNegativeFloat = Field(
        ..., example=0.5, description="Petal length in cm"
    )
    petal_width: NonNegativeFloat = Field(
        ..., example=0.5, description="Petal width in cm"
    )


class IrisPrediction(BaseModel):
    species: str = Field(..., example="setosa", description="Predicted species")

from fastkafka import FastKafka

kafka_brokers = {
    "localhost": {
        "url": "localhost",
        "description": "local development kafka broker",
        "port": 9092,
    },
    "production": {
        "url": "kafka.airt.ai",
        "description": "production kafka broker",
        "port": 9092,
        "protocol": "kafka-secure",
        "security": {"type": "plain"},
    },
}

kafka_app = FastKafka(
    title="Iris predictions",
    kafka_brokers=kafka_brokers,
    lifespan=lifespan,
)

@kafka_app.consumes(topic="input_data", auto_offset_reset="latest")
async def on_input_data(msg: IrisInputData):
    species_class = ml_models["iris_predictor"].predict(
        [[msg.sepal_length, msg.sepal_width, msg.petal_length, msg.petal_width]]
    )[0]

    await to_predictions(species_class)


@kafka_app.produces(topic="predictions")
async def to_predictions(species_class: int) -> IrisPrediction:
    iris_species = ["setosa", "versicolor", "virginica"]

    prediction = IrisPrediction(species=iris_species[species_class])
    return prediction

Creating requirements.txt file

The above code only requires FastKafka. So, we will add only that to the requirements.txt file, but you can add additional requirements to it as well.

fastkafka>=0.3.0

Here we are using requirements.txt to store the project’s dependencies. However, other methods like setup.py, pipenv, and wheel files can also be used. setup.py is commonly used for packaging and distributing Python modules, while pipenv is a tool used for managing virtual environments and package dependencies. wheel files are built distributions of Python packages that can be installed with pip.

Creating Dockerfile

# (1)
FROM python:3.9-slim-bullseye
# (2)
WORKDIR /project
# (3)
COPY application.py requirements.txt /project/
# (4)
RUN pip install --no-cache-dir --upgrade -r /project/requirements.txt
# (5)
CMD ["fastkafka", "run", "--num-workers", "2", "--kafka-broker", "production", "application:kafka_app"]

1. Start from the official Python base image.

2. Set the current working directory to /project.

   This is where we’ll put the requirements.txt file and the application.py file.

3. Copy the application.py file and requirements.txt file inside the /project directory.

4. Install the package dependencies in the requirements file.

   The --no-cache-dir option tells pip to not save the downloaded packages locally, as that is only if pip was going to be run again to install the same packages, but that’s not the case when working with containers.

   The --upgrade option tells pip to upgrade the packages if they are already installed.

5. Set the command to run the fastkafka run command.

   CMD takes a list of strings, each of these strings is what you would type in the command line separated by spaces.

   This command will be run from the current working directory, the same /project directory you set above with WORKDIR /project.

   We supply additional parameters --num-workers and --kafka-broker for the run command. Finally, we specify the location of our FastKafka application as a command argument.

   To learn more about fastkafka run command please check the CLI docs.

Build the Docker Image

Now that all the files are in place, let’s build the container image.

1. Go to the project directory (where your Dockerfile is, containing your application.py file).

2. Run the following command to build the image:

   docker build -t fastkafka_project_image .

   This command will create a docker image with the name fastkafka_project_image and the latest tag.

That’s it! You have now built a docker image for your FastKafka project.

Start the Docker Container

Run a container based on the built image:

docker run -d --name fastkafka_project_container fastkafka_project_image

Additional Security

Trivy is an open-source tool that scans Docker images for vulnerabilities. It can be integrated into your CI/CD pipeline to ensure that your images are secure and free from known vulnerabilities. Here’s how you can use trivy to scan your fastkafka_project_image:

1. Install trivy on your local machine by following the instructions provided in the official trivy documentation.

2. Run the following command to scan your fastkafka_project_image:

   trivy image fastkafka_project_image

   This command will scan your fastkafka_project_image for any vulnerabilities and provide you with a report of its findings.

3. Fix any vulnerabilities identified by trivy. You can do this by updating the vulnerable package to a more secure version or by using a different package altogether.

4. Rebuild your fastkafka_project_image and repeat steps 2 and 3 until trivy reports no vulnerabilities.

By using trivy to scan your Docker images, you can ensure that your containers are secure and free from known vulnerabilities.

Example repo

A FastKafka based library which uses above mentioned Dockerfile to build a docker image can be found here