Important

You are viewing documentation for an older version of Confluent Platform. For the latest, click here.

Azure Functions Sink Connector for Confluent Platform

The Kafka Connect Azure Functions Sink Connector integrates Kafka with Azure Functions.

Note

If you are using Confluent Cloud, see https://docs.confluent.io/cloud/current/connectors/cc-azure-functions-sink.html for the cloud Quick Start.

The connector consumes records from Kafka topic(s) and executes an Azure Function. Each request sent to Azure Functions contains can contain up to max.batch.size records. The connector can also send many requests concurrently, set by max.outstanding.requests.

The target function must be configured and ready to accept requests with the following JSON format:

[
  {
    "key": ...,
    "value": ...,
    "topic": string,
    "partition": <number>,
    "offset": <number>,
    "timestamp": <number>
  },
  ...,
]

Note

The key and value are encoded as follows:

  • String, int, long, float, double, boolean, null are encoded as-is into JSON.
  • Structs are converted to JSON and exported without the schema.
  • byte[] is encoded as a base64 String and sent as a JSON string.
  • Any other Java objects are converted to String using toString(), and then sent as JSON strings.

The connector receives the responses from the Azure Function and writes it to a result or error topic (set by configurations) depending on the HTTP Response code. Response code 400 and above are considered errors and anything below is a success.

The connector attempts to map each response to a single record before producing it to the corresponding topic. It can receive the responses from the Azure Function in the following three formats.

  1. The first format is JSON:

    [
  {
    "payload": {
      "result": ...,
      "topic": string,
      "partition": <number>,
      "offset": <number>,
    }
  },
  ...
]

    This list can be out of order relative to the order that the records were provided. The connector will correctly match the record to the result based off its Kafka coordinates. However the list must be one-to-one to the list of records that were sent in the request.

  2. The second format is a JSON list:

    [
  ...,
  ...,
  ...
]

    As long as the list is one-to-one to the list of records, the list will be assumed to be ordered and matched with the corresponding records.

  3. The third format can be any format that does not satisfy either of the above formats. The connector will report the entire response for each individual record (one-to-many correlation).

Install the Azure Functions Connector

You can install this connector by using the Confluent Hub client (recommended) or you can manually download the ZIP file.

Install the connector using Confluent Hub

Prerequisite
Confluent Hub Client must be installed. This is installed by default with Confluent Enterprise.

Navigate to your Confluent Platform installation directory and run the following command to install the latest (latest) connector version. The connector must be installed on every machine where Connect will run.

confluent-hub install confluentinc/kafka-connect-azure-functions:latest

You can install a specific version by replacing latest with a version number. For example:

confluent-hub install confluentinc/kafka-connect-azure-functions:1.0.0-preview

Install the connector manually

Download and extract the ZIP file for your connector and then follow the manual connector installation instructions.

License

You can use this connector for a 30-day trial period without a license key.

After 30 days, this connector is available under a Confluent enterprise license. Confluent issues enterprise license keys to subscribers, along with providing enterprise-level support for Confluent Platform and your connectors. If you are a subscriber, please contact Confluent Support at support@confluent.io for more information.

See Confluent Platform license for license properties and License topic configuration for information about the license topic.

Configuration Properties

For a complete list of configuration properties for this connector, see Azure Functions Sink Connector Configuration Properties.

Note

For an example of how to get Kafka Connect connected to Confluent Cloud, see Distributed Cluster in Connect Kafka Connect to Confluent Cloud.

Quick Start

This quick start uses the Azure Functions Sink Connector to consume records and execute an example Azure Function.

Additional examples can be found in the Feature Descriptions and Examples section below.

Prerequisites

  1. Before starting the connector, create and deploy an Azure Functions instance.

    Important

    Make sure to select the Node.js runtime stack and to create the function in-portal.

    • Copy and paste this code into the index.js file in the portal.

      module.exports = async function (context, req) {
    context.log('JavaScript HTTP trigger function processed a request.');
    context.res = {
        status: 200,
        body: req.body
    };
};

    • Copy the function url from the portal and save it for later. Azure Functions should now be set up for the connector.

      Note

      This example function just mirrors the request and sends the request body back in the response.

  2. Install the connector through the Confluent Hub Client.

    # run from your CP installation directory
confluent-hub install confluentinc/kafka-connect-azure-functions:latest

  3. Start Confluent Platform using the Confluent CLI commands.

    Tip

    The command syntax for the Confluent CLI development commands changed in 5.3.0. These commands have been moved to confluent local. For example, the syntax for confluent start is now confluent local start. For more information, see confluent local.

    confluent local start

  4. Produce test data to the functions-test topic in Kafka using the Confluent CLI confluent local produce command.

    echo key1,value1 | confluent local produce functions-test -- --property parse.key=true --property key.separator=,
echo key2,value2 | confluent local produce functions-test -- --property parse.key=true --property key.separator=,
echo key3,value3 | confluent local produce functions-test -- --property parse.key=true --property key.separator=,

  5. Create a azure-functions-sink.json file with the following contents:

    {
  "name": "azure-functions",
  "config": {
    "topics": "functions-test",
    "tasks.max": "1",
    "connector.class": "io.confluent.connect.azure.functions.AzureFunctionsSinkConnector",
    "key.converter":"org.apache.kafka.connect.storage.StringConverter",
    "value.converter":"org.apache.kafka.connect.storage.StringConverter",
    "confluent.topic.bootstrap.servers": "localhost:9092",
    "confluent.topic.replication.factor": "1",
    "function.url": "<the copied function url>",
    "reporter.bootstrap.servers": "localhost:9092",
    "reporter.error.topic.name": "test-error",
    "reporter.error.topic.replication.factor": 1,
    "reporter.error.topic.key.format": "string",
    "reporter.error.topic.value.format": "string",
    "reporter.result.topic.name": "test-result",
    "reporter.result.topic.key.format": "string",
    "reporter.result.topic.value.format": "string",
    "reporter.result.topic.replication.factor": 1
  }
}

    Caution

    Do not forget to change the function.url value in the JSON file to the copied function url.

  6. Load the Azure Functions Sink Connector.

    Caution

    You must include a double dash (--) between the topic name and your flag. For more information, see this post.

    confluent local load azure-functions -- -d path/to/azure-functions-sink.json

    Important

    Don’t use the Confluent CLI commands in production environments.

  7. Confirm that the connector is in a RUNNING state.

    confluent local status azure-functions

  8. Confirm that the messages were delivered to the result topic in Kafka

    confluent local consume test-result -- --from-beginning

  9. Cleanup resources

    • Delete the connector

      confluent local unload azure-functions

    • Stop Confluent Platform

      confluent local stop

    • Delete the created Azure Function in the Azure portal.