Kafka Streaming

This is a collection of data movement capabilities. This streaming job copies data from Kafka to Iceberg.

Table of Contents

• Deserialization
• Job creation
• Tests

Deserialization

Currently, two deserialization format supported.

1. JSON
2. AVRO

JSON

In the Spark configuration, a user-defined reference json schema can be defined,
and the system processes the binary data accordingly. Otherwise,
It considers the schema of the first row and assumes the rest of the rows is compatible.

Avro

Converts binary data according to the schema defined by the user or retrieves the schema from the schema registry.

Job creation

• Go to Spark Jobs.
• Click on Create New.

Specify the following parameters (these are examples, you can change them based on your preference):

• Name: kafka-streaming-job
• Docker Image: iomete/iomete_kafka_streaming_job:0.2.1
• Main application file: local:///app/driver.py
• Environment Variables: LOG_LEVEL: INFO or ERROR
• Config file:

{
  kafka: {
      bootstrap_servers: "localhost:9092",
      topic: "usage.spark.0",
      serialization_format: json,
      group_id: group_1,
      starting_offsets: latest,
      trigger: {
        interval: 5
        unit: seconds # minutes
      },
      schema_registry_url: "http://127.0.0.1:8081"
  },
  database: {
    schema: default,
    table: spark_usage_20
  }
}

Configuration properties

Create Spark Job - Deployment

Create Spark Job - Instance

You can use Environment Variables to store your sensitive data like password, secrets, etc. Then you can use these variables in your config file using the ${ENV_NAME} syntax.

Create Spark Job - Application Configurations

Create Spark Job - Application dependencies

Tests

Prepare the dev environment

virtualenv .env #or python3 -m venv .env
source .env/bin/activate

pip install -e ."[dev]"

Run test

python3 -m pytest