FastKafka

Effortless Kafka integration for your web services

---

---

FastKafka is a powerful and easy-to-use Python library for building asynchronous services that interact with Kafka topics. Built on top of Pydantic, AIOKafka and AsyncAPI, FastKafka simplifies the process of writing producers and consumers for Kafka topics, handling all the parsing, networking, task scheduling and data generation automatically. With FastKafka, you can quickly prototype and develop high-performance Kafka-based services with minimal code, making it an ideal choice for developers looking to streamline their workflow and accelerate their projects.

---

⭐⭐⭐ Stay in touch ⭐⭐⭐

Please show your support and stay in touch by:

• giving our GitHub repository a star, and

• joining our Discord server.

Your support helps us to stay in touch with you and encourages us to continue developing and improving the library. Thank you for your support!

---

🐝🐝🐝 We were busy lately 🐝🐝🐝

Install

FastKafka works on macOS, Linux, and most Unix-style operating systems. You can install base version of fastkafka with pip as usual:

pip install fastkafka

To install fastkafka with testing features please use:

pip install fastkafka[test]

To install fastkafka with asyncapi docs please use:

pip install fastkafka[docs]

To install fastkafka with all the features please use:

pip install fastkafka[test,docs]

Tutorial

You can start an interactive tutorial in Google Colab by clicking the button below:

Writing server code

Here is an example python script using FastKafka that takes data from a Kafka topic, makes a prediction using a predictive model, and outputs the prediction to another Kafka topic.

Preparing the demo model

First we will prepare our model using the Iris dataset so that we can demonstrate the predictions using FastKafka. The following call downloads the dataset and trains the model.

We will wrap the model creation into a lifespan of our app so that the model is created just before the app is started.

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()

Messages

FastKafka uses Pydantic to parse input JSON-encoded data into Python objects, making it easy to work with structured data in your Kafka-based applications. Pydantic’s BaseModel class allows you to define messages using a declarative syntax, making it easy to specify the fields and types of your messages.

This example defines two message classes for use in a FastKafka application:

• The IrisInputData class is used to represent input data for a predictive model. It has four fields of type NonNegativeFloat, which is a subclass of float that only allows non-negative floating point values.

• The IrisPrediction class is used to represent the output of the predictive model. It has a single field species of type string representing the predicted species.

These message classes will be used to parse and validate incoming data in Kafka consumers and producers.

from pydantic import BaseModel, Field, NonNegativeFloat


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")

Application

This example shows how to initialize a FastKafka application.

It starts by defining a dictionary called kafka_brokers, which contains two entries: "localhost" and "production", specifying local development and production Kafka brokers. Each entry specifies the URL, port, and other details of a Kafka broker. This dictionary is used for both generating the documentation and later to run the actual server against one of the given kafka broker.

Next, an object of the FastKafka class is initialized with the minimum set of arguments:

• kafka_brokers: a dictionary used for generation of documentation

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,
)

Function decorators

FastKafka provides convenient function decorators @kafka_app.consumes and @kafka_app.produces to allow you to delegate the actual process of

• consuming and producing data to Kafka, and

• decoding and encoding JSON encode messages

from user defined functions to the framework. The FastKafka framework delegates these jobs to AIOKafka and Pydantic libraries.

These decorators make it easy to specify the processing logic for your Kafka consumers and producers, allowing you to focus on the core business logic of your application without worrying about the underlying Kafka integration.

This following example shows how to use the @kafka_app.consumes and @kafka_app.produces decorators in a FastKafka application:

• The @kafka_app.consumes decorator is applied to the on_input_data function, which specifies that this function should be called whenever a message is received on the “input_data" Kafka topic. The on_input_data function takes a single argument which is expected to be an instance of the IrisInputData message class. Specifying the type of the single argument is instructing the Pydantic to use IrisInputData.parse_raw() on the consumed message before passing it to the user defined function on_input_data.

• The @produces decorator is applied to the to_predictions function, which specifies that this function should produce a message to the “predictions" Kafka topic whenever it is called. The to_predictions function takes a single integer argument species_class representing one of three possible strign values predicted by the mdoel. It creates a new IrisPrediction message using this value and then returns it. The framework will call the IrisPrediction.json().encode("utf-8") function on the returned value and produce it to the specified topic.

@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]

    to_predictions(species_class)


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

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

Testing the service

The service can be tested using the Tester instances which internally starts InMemory implementation of Kafka broker.

The Tester will redirect your consumes and produces decorated functions to the InMemory Kafka broker so that you can quickly test your app without the need for a running Kafka broker and all its dependencies.

from fastkafka.testing import Tester

msg = IrisInputData(
    sepal_length=0.1,
    sepal_width=0.2,
    petal_length=0.3,
    petal_width=0.4,
)

# Start Tester app and create InMemory Kafka broker for testing
async with Tester(kafka_app) as tester:
    # Send IrisInputData message to input_data topic
    await tester.to_input_data(msg)

    # Assert that the kafka_app responded with IrisPrediction in predictions topic
    await tester.awaited_mocks.on_predictions.assert_awaited_with(
        IrisPrediction(species="setosa"), timeout=2
    )

[INFO] fastkafka._testing.in_memory_broker: InMemoryBroker._start() called
[INFO] fastkafka._testing.in_memory_broker: InMemoryBroker._patch_consumers_and_producers(): Patching consumers and producers!
[INFO] fastkafka._testing.in_memory_broker: InMemoryBroker starting
[INFO] fastkafka._application.app: _create_producer() : created producer using the config: '{'bootstrap_servers': 'localhost:9092'}'
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaProducer patched start() called()
[INFO] fastkafka._application.app: _create_producer() : created producer using the config: '{'bootstrap_servers': 'localhost:9092'}'
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaProducer patched start() called()
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() starting...
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer created using the following parameters: {'bootstrap_servers': 'localhost:9092', 'auto_offset_reset': 'latest', 'max_poll_records': 100}
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched start() called()
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer started.
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched subscribe() called
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer.subscribe(), subscribing to: ['input_data']
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer subscribed.
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() starting...
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer created using the following parameters: {'bootstrap_servers': 'localhost:9092', 'auto_offset_reset': 'earliest', 'max_poll_records': 100}
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched start() called()
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer started.
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched subscribe() called
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer.subscribe(), subscribing to: ['predictions']
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer subscribed.
[INFO] fastkafka._components.aiokafka_consumer_loop: _aiokafka_consumer_loop(): Consumer loop shutting down, waiting for send_stream to drain...
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched stop() called
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer stopped.
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() finished.
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaProducer patched stop() called
[INFO] fastkafka._components.aiokafka_consumer_loop: _aiokafka_consumer_loop(): Consumer loop shutting down, waiting for send_stream to drain...
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaConsumer patched stop() called
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer stopped.
[INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() finished.
[INFO] fastkafka._testing.in_memory_broker: AIOKafkaProducer patched stop() called
[INFO] fastkafka._testing.in_memory_broker: InMemoryBroker._stop() called
[INFO] fastkafka._testing.in_memory_broker: InMemoryBroker stopping

Recap

We have created a Iris classification model and encapulated it into our fastkafka application. The app will consume the IrisInputData from the input_data topic and produce the predictions to predictions topic.

To test the app we have:

1. Created the app

2. Started our Tester class which mirrors the developed app topics for testing purposes

3. Sent IrisInputData message to input_data topic

4. Asserted and checked that the developed iris classification service has reacted to IrisInputData message

Running the service

The service can be started using builtin faskafka run CLI command. Before we can do that, we will concatenate the code snippets from above and save them in a file "application.py"

# 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]

    to_predictions(species_class)


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

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

To run the service, you will need a running Kafka broker on localhost as specified in the kafka_brokers parameter above. We can start the Kafka broker locally using the ApacheKafkaBroker.

To use ApacheKafkaBroker, you need to install JRE and Kafka to your environment. To simplify this process, fastkafka comes with a CLI command that does just that, to run it, in your terminal execute the following:

fastkafka testing install_deps

Now we can run ApacheKafkaBroker that will start a Kafka broker instance for us.

from fastkafka.testing import ApacheKafkaBroker

broker = ApacheKafkaBroker(apply_nest_asyncio=True)

broker.start()

[INFO] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.start(): entering...
[WARNING] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.start(): (<_UnixSelectorEventLoop running=True closed=False debug=False>) is already running!
[WARNING] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.start(): calling nest_asyncio.apply()
[INFO] fastkafka._components.test_dependencies: Java is already installed.
[INFO] fastkafka._components.test_dependencies: But not exported to PATH, exporting...
[INFO] fastkafka._components.test_dependencies: Kafka is installed.
[INFO] fastkafka._components.test_dependencies: But not exported to PATH, exporting...
[INFO] fastkafka._testing.apache_kafka_broker: Starting zookeeper...
[INFO] fastkafka._testing.apache_kafka_broker: Starting kafka...
[INFO] fastkafka._testing.apache_kafka_broker: Local Kafka broker up and running on 127.0.0.1:9092
[INFO] fastkafka._testing.apache_kafka_broker: <class 'fastkafka.testing.ApacheKafkaBroker'>.start(): returning 127.0.0.1:9092
[INFO] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.start(): exited.

'127.0.0.1:9092'

Then, we start the FastKafka service by running the following command in the folder where the application.py file is located:

fastkafka run --num-workers=2 --kafka-broker localhost application:kafka_app

In the above command, we use --num-workers option to specify how many workers to launch and we use --kafka-broker option to specify which kafka broker configuration to use from earlier specified kafka_brokers

[801767]: [INFO] fastkafka._application.app: set_kafka_broker() : Setting bootstrap_servers value to 'localhost:9092'
[801765]: [INFO] fastkafka._application.app: set_kafka_broker() : Setting bootstrap_servers value to 'localhost:9092'
[801767]: [INFO] fastkafka._application.app: _create_producer() : created producer using the config: '{'bootstrap_servers': 'localhost:9092'}'
[801765]: [INFO] fastkafka._application.app: _create_producer() : created producer using the config: '{'bootstrap_servers': 'localhost:9092'}'
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() starting...
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() starting...
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer created using the following parameters: {'bootstrap_servers': 'localhost:9092', 'auto_offset_reset': 'latest', 'max_poll_records': 100}
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer created using the following parameters: {'bootstrap_servers': 'localhost:9092', 'auto_offset_reset': 'latest', 'max_poll_records': 100}
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer started.
[801767]: [INFO] aiokafka.consumer.subscription_state: Updating subscribed topics to: frozenset({'input_data'})
[801767]: [INFO] aiokafka.consumer.consumer: Subscribed to topic(s): {'input_data'}
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer subscribed.
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer started.
[801765]: [INFO] aiokafka.consumer.subscription_state: Updating subscribed topics to: frozenset({'input_data'})
[801765]: [INFO] aiokafka.consumer.consumer: Subscribed to topic(s): {'input_data'}
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer subscribed.
[801765]: [ERROR] aiokafka.cluster: Topic input_data not found in cluster metadata
[801765]: [INFO] aiokafka.consumer.group_coordinator: Metadata for topic has changed from {} to {'input_data': 0}. 
[801767]: [WARNING] aiokafka.cluster: Topic input_data is not available during auto-create initialization
[801767]: [INFO] aiokafka.consumer.group_coordinator: Metadata for topic has changed from {} to {'input_data': 0}. 
[801767]: [ERROR] aiokafka: Unable connect to node with id 0: [Errno 111] Connect call failed ('192.168.112.2', 9092)
[801765]: [ERROR] aiokafka: Unable connect to node with id 0: [Errno 111] Connect call failed ('192.168.112.2', 9092)
[801767]: [ERROR] aiokafka: Unable to update metadata from [0]
[801765]: [ERROR] aiokafka: Unable to update metadata from [0]
^C
Starting process cleanup, this may take a few seconds...
[INFO] fastkafka._server: terminate_asyncio_process(): Terminating the process 801765...
[INFO] fastkafka._server: terminate_asyncio_process(): Terminating the process 801767...
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: _aiokafka_consumer_loop(): Consumer loop shutting down, waiting for send_stream to drain...
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer stopped.
[801765]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() finished.
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: _aiokafka_consumer_loop(): Consumer loop shutting down, waiting for send_stream to drain...
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop(): Consumer stopped.
[801767]: [INFO] fastkafka._components.aiokafka_consumer_loop: aiokafka_consumer_loop() finished.

You need to interupt running of the cell above by selecting Runtime->Interupt execution on the toolbar above.

Finally, we can stop the local Kafka Broker:

broker.stop()

[INFO] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.stop(): entering...
[INFO] fastkafka._components._subprocess: terminate_asyncio_process(): Terminating the process 801303...
[INFO] fastkafka._components._subprocess: terminate_asyncio_process(): Process 801303 was already terminated.
[INFO] fastkafka._components._subprocess: terminate_asyncio_process(): Terminating the process 800930...
[INFO] fastkafka._components._subprocess: terminate_asyncio_process(): Process 800930 was already terminated.
[INFO] fastkafka._testing.apache_kafka_broker: ApacheKafkaBroker.stop(): exited.

Documentation

The kafka app comes with builtin documentation generation using AsyncApi HTML generator.

AsyncApi requires Node.js to be installed and we provide the following convenience command line for it:

fastkafka docs install_deps

[INFO] fastkafka._components.docs_dependencies: AsyncAPI generator installed

To generate the documentation programatically you just need to call the folloving command:

fastkafka docs generate application:kafka_app

[INFO] fastkafka._components.asyncapi: New async specifications generated at: '/work/fastkafka/nbs/asyncapi/spec/asyncapi.yml'
[INFO] fastkafka._components.asyncapi: Async docs generated at 'asyncapi/docs'
[INFO] fastkafka._components.asyncapi: Output of '$ npx -y -p @asyncapi/generator ag asyncapi/spec/asyncapi.yml @asyncapi/html-template -o asyncapi/docs --force-write'

Done! ✨
Check out your shiny new generated files at /work/fastkafka/nbs/asyncapi/docs.

. This will generate the asyncapi folder in relative path where all your documentation will be saved. You can check out the content of it with:

ls -l asyncapi

total 8
drwxrwxr-x 4 kumaran kumaran 4096 Mar 21 09:14 docs
drwxrwxr-x 2 kumaran kumaran 4096 Mar 21 09:14 spec

In docs folder you will find the servable static html file of your documentation. This can also be served using our fastkafka docs serve CLI command (more on that in our guides).

In spec folder you will find a asyncapi.yml file containing the async API specification of your application.

We can locally preview the generated documentation by running the following command:

fastkafka docs serve application:kafka_app

[INFO] fastkafka._components.asyncapi: New async specifications generated at: '/work/fastkafka/nbs/asyncapi/spec/asyncapi.yml'
[INFO] fastkafka._components.asyncapi: Async docs generated at 'asyncapi/docs'
[INFO] fastkafka._components.asyncapi: Output of '$ npx -y -p @asyncapi/generator ag asyncapi/spec/asyncapi.yml @asyncapi/html-template -o asyncapi/docs --force-write'

Done! ✨
Check out your shiny new generated files at /work/fastkafka/nbs/asyncapi/docs.


Serving documentation on http://127.0.0.1:8000
^C
Interupting serving of documentation and cleaning up...

From the parameters passed to the application constructor, we get the documentation bellow:

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,
)

The following documentation snippet are for the consumer as specified in the code above:

The following documentation snippet are for the producer as specified in the code above:

Finally, all messages as defined as subclasses of BaseModel are documented as well:

License

FastKafka is licensed under the Apache License 2.0

A permissive license whose main conditions require preservation of copyright and license notices. Contributors provide an express grant of patent rights. Licensed works, modifications, and larger works may be distributed under different terms and without source code.

The full text of the license can be found here.