Skip to content

kafka.md

Latest commit

 

History

History
1001 lines (796 loc) · 44.6 KB

kafka.md

File metadata and controls

1001 lines (796 loc) · 44.6 KB
title Kafka Consumer
description Utility
status new

The Kafka utility transparently handles message deserialization, provides an intuitive developer experience, and integrates seamlessly with the rest of the Powertools for AWS Lambda ecosystem.

flowchart LR
    KafkaTopic["Kafka Topic"] --> MSK["Amazon MSK"]
    KafkaTopic --> MSKServerless["Amazon MSK Serverless"]
    KafkaTopic --> SelfHosted["Self-hosted Kafka"]
    MSK --> EventSourceMapping["Event Source Mapping"]
    MSKServerless --> EventSourceMapping
    SelfHosted --> EventSourceMapping
    EventSourceMapping --> Lambda["Lambda Function"]
    Lambda --> KafkaUtility["Kafka Utility"]
    KafkaUtility --> Deserialization["Deserialization"]
    Deserialization --> YourLogic["Your Business Logic"]
Loading

Key features

  • Automatic deserialization of Kafka messages (JSON, Avro, and Protocol Buffers)
  • Simplified event record handling with familiar Kafka ConsumerRecords interface
  • Support for key and value deserialization
  • Support for ESM with and without Schema Registry integration
  • Proper error handling for deserialization issues

Terminology

Event Source Mapping (ESM) A Lambda feature that reads from streaming sources (like Kafka) and invokes your Lambda function. It manages polling, batching, and error handling automatically, eliminating the need for consumer management code.

Record Key and Value A Kafka messages contain two important parts: an optional key that determines the partition and a value containing the actual message data. Both are base64-encoded in Lambda events and can be independently deserialized.

Deserialization Is the process of converting binary data (base64-encoded in Lambda events) into usable Java objects according to a specific format like JSON, Avro, or Protocol Buffers. Powertools handles this conversion automatically.

DeserializationType enum Contains parameters that tell Powertools how to interpret message data, including the format type (JSON, Avro, Protocol Buffers).

Schema Registry Is a centralized service that stores and validates schemas, ensuring producers and consumers maintain compatibility when message formats evolve over time.

Moving from traditional Kafka consumers

Lambda processes Kafka messages as discrete events rather than continuous streams, requiring a different approach to consumer development that Powertools for AWS helps standardize.

Aspect Traditional Kafka Consumers Lambda Kafka Consumer
Model Pull-based (you poll for messages) Push-based (Lambda invoked with messages)
Scaling Manual scaling configuration Automatic scaling to partition count
State Long-running application with state Stateless, ephemeral executions
Offsets Manual offset management Automatic offset commitment
Schema Validation Client-side schema validation Optional Schema Registry integration with Event Source Mapping
Error Handling Per-message retry control Batch-level retry policies

Getting started

Installation

Add the Powertools for AWS Lambda Kafka dependency to your project. Make sure to also add the kafka-clients library as a dependency. The utility supports kafka-clients >= 3.0.0.

=== "Maven"

```xml
<dependency>
    <groupId>software.amazon.lambda</groupId>
    <artifactId>powertools-kafka</artifactId>
    <version>{{ powertools.version }}</version>
</dependency>
<!-- Kafka clients dependency - compatibility works for >= 3.0.0 -->
<dependency>
    <groupId>org.apache.kafka</groupId>
    <artifactId>kafka-clients</artifactId>
    <version>4.0.0</version>
</dependency>
```

=== "Gradle"

```gradle
dependencies {
    implementation 'software.amazon.lambda:powertools-kafka:{{ powertools.version }}'
    // Kafka clients dependency - compatibility works for >= 3.0.0
    implementation 'org.apache.kafka:kafka-clients:4.0.0'
}
```

Required resources

To use the Kafka utility, you need an AWS Lambda function configured with a Kafka event source. This can be Amazon MSK, MSK Serverless, or a self-hosted Kafka cluster.

=== "getting_started_with_msk.yaml"

```yaml
AWSTemplateFormatVersion: '2010-09-09'
Transform: AWS::Serverless-2016-10-31
Resources:
  KafkaConsumerFunction:
    Type: AWS::Serverless::Function
    Properties:
      Handler: org.example.KafkaHandler::handleRequest
      Runtime: java21
      Timeout: 30
      Events:
        MSKEvent:
          Type: MSK
          Properties:
            StartingPosition: LATEST
            Stream: !GetAtt MyMSKCluster.Arn
            Topics:
              - my-topic-1
              - my-topic-2
      Policies:
        - AWSLambdaMSKExecutionRole
```

Using ESM with Schema Registry

The Event Source Mapping configuration determines which mode is used. With JSON, Lambda converts all messages to JSON before invoking your function. With SOURCE mode, Lambda preserves the original format, requiring you function to handle the appropriate deserialization.

Powertools for AWS supports both Schema Registry integration modes in your Event Source Mapping configuration.

Processing Kafka events

The Kafka utility transforms raw Lambda Kafka events into an intuitive format for processing. To handle messages effectively, you'll need to configure the @Deserialization annotation that matches your data format. Based on the deserializer you choose, incoming records are directly transformed into your business objects which can be auto-generated classes from Avro / Protobuf or simple POJOs.

???+ tip "Using Avro is recommended" We recommend Avro for production Kafka implementations due to its schema evolution capabilities, compact binary format, and integration with Schema Registry. This offers better type safety and forward/backward compatibility compared to JSON.

=== "Avro Messages"

```java hl_lines="18 21"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class AvroKafkaHandler implements RequestHandler<ConsumerRecords<String, User>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(AvroKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_AVRO)
    public String handleRequest(ConsumerRecords<String, User> records, Context context) {
        for (ConsumerRecord<String, User> record : records) {
            User user = record.value(); // User class is auto-generated from Avro schema
            LOGGER.info("Processing user: {}, age {}", user.getName(), user.getAge());
        }
        return "OK";
    }
}
```

=== "Protocol Buffers"

```java hl_lines="18 21"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class ProtobufKafkaHandler implements RequestHandler<ConsumerRecords<String, UserProto.User>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(ProtobufKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_PROTOBUF)
    public String handleRequest(ConsumerRecords<String, UserProto.User> records, Context context) {
        for (ConsumerRecord<String, UserProto.User> record : records) {
            UserProto.User user = record.value(); // UserProto.User class is auto-generated from Protocol Buffer schema
            LOGGER.info("Processing user: {}, age {}", user.getName(), user.getAge());
        }
        return "OK";
    }
}
```

=== "JSON Messages"

```java hl_lines="18 21"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class JsonKafkaHandler implements RequestHandler<ConsumerRecords<String, User>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(JsonKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<String, User> records, Context context) {
        for (ConsumerRecord<String, User> record : records) {
            User user = record.value(); // Deserialized JSON object into User POJO
            LOGGER.info("Processing user: {}, age {}", user.getName(), user.getAge());
        }
        return "OK";
    }
}
```

???+ tip "Full examples on GitHub" A full example including how to generate Avro and Protobuf Java classes can be found on GitHub at https://github.com/aws-powertools/powertools-lambda-java/tree/main/examples/powertools-examples-kafka.

Deserializing keys and values

The @Deserialization annotation deserializes both keys and values based on your type configuration. This flexibility allows you to work with different data formats in the same message.

=== "Key and Value Deserialization"

```java hl_lines="22"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class KeyValueKafkaHandler implements RequestHandler<ConsumerRecords<ProductKey, ProductInfo>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(KeyValueKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_AVRO)
    public String handleRequest(ConsumerRecords<ProductKey, ProductInfo> records, Context context) {
        for (ConsumerRecord<ProductKey, ProductInfo> record : records) {
            // Access both deserialized components
            ProductKey key = record.key(); // ProductKey class is auto-generated from Avro schema
            ProductInfo product = record.value(); // ProductInfo class is auto-generated from Avro schema

            LOGGER.info("Processing product ID: {}", key.getProductId());
            LOGGER.info("Product: {} - ${}", product.getName(), product.getPrice());
        }
        return "OK";
    }
}
```

=== "Value-Only Deserialization"

```java hl_lines="22"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class ValueOnlyKafkaHandler implements RequestHandler<ConsumerRecords<String, Order>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(ValueOnlyKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<String, Order> records, Context context) {
        for (ConsumerRecord<String, Order> record : records) {
            // Key remains as string (if present)
            String key = record.key();
            if (key != null) {
                LOGGER.info("Message key: {}", key);
            }

            // Value is deserialized as JSON
            Order order = record.value();
            LOGGER.info("Order #{} - Total: ${}", order.getOrderId(), order.getTotal());
        }
        return "OK";
    }
}
```

Handling primitive types

When working with primitive data types (strings, integers, etc.) rather than structured objects, you can use any deserialization type such as KAFKA_JSON. Simply place the primitive type like Integer or String in the ConsumerRecords generic type parameters, and the library will automatically handle primitive type deserialization.

???+ tip "Common pattern: Keys with primitive values" Using primitive types (strings, integers) as Kafka message keys is a common pattern for partitioning and identifying messages. Powertools automatically handles these primitive keys without requiring special configuration, making it easy to implement this popular design pattern.

=== "Primitive key"

```java hl_lines="18 22"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class PrimitiveKeyHandler implements RequestHandler<ConsumerRecords<Integer, Customer>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(PrimitiveKeyHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<Integer, Customer> records, Context context) {
        for (ConsumerRecord<Integer, Customer> record : records) {
            // Key is automatically deserialized as Integer
            Integer key = record.key();

            // Value is deserialized as JSON
            Customer customer = record.value();

            LOGGER.info("Key: {}", key);
            LOGGER.info("Name: {}", customer.getName());
            LOGGER.info("Email: {}", customer.getEmail());
        }
        return "OK";
    }
}
```

=== "Primitive key and value"

```java hl_lines="18 22"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class PrimitiveHandler implements RequestHandler<ConsumerRecords<String, String>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(PrimitiveHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<String, String> records, Context context) {
        for (ConsumerRecord<String, String> record : records) {
            // Key is automatically deserialized as String
            String key = record.key();

            // Value is automatically deserialized as String
            String value = record.value();

            LOGGER.info("Key: {}", key);
            LOGGER.info("Value: {}", value);
        }
        return "OK";
    }
}
```

Message format support and comparison

The Kafka utility supports multiple serialization formats to match your existing Kafka implementation. Choose the format that best suits your needs based on performance, schema evolution requirements, and ecosystem compatibility.

???+ tip "Selecting the right format" For new applications, consider Avro or Protocol Buffers over JSON. Both provide schema validation, evolution support, and significantly better performance with smaller message sizes. Avro is particularly well-suited for Kafka due to its built-in schema evolution capabilities.

=== "Supported Formats"

| Format | DeserializationType | Description | Required Dependencies |
|--------|---------------------|-------------|----------------------|
| **JSON** | `KAFKA_JSON` | Human-readable text format | Jackson |
| **Avro** | `KAFKA_AVRO` | Compact binary format with schema | Apache Avro |
| **Protocol Buffers** | `KAFKA_PROTOBUF` | Efficient binary format | Protocol Buffers |
| **Lambda Default** | `LAMBDA_DEFAULT` | Uses Lambda's built-in deserialization (equivalent to removing the `@Deserialization` annotation) | None |

=== "Format Comparison"

| Feature | JSON | Avro | Protocol Buffers |
|---------|------|------|-----------------|
| **Schema Definition** | Optional | Required schema file | Required .proto file |
| **Schema Evolution** | None | Strong support | Strong support |
| **Size Efficiency** | Low | High | Highest |
| **Processing Speed** | Slower | Fast | Fastest |
| **Human Readability** | High | Low | Low |
| **Implementation Complexity** | Low | Medium | Medium |
| **Additional Dependencies** | None | Apache Avro | Protocol Buffers |

Choose the serialization format that best fits your needs:

  • JSON: Best for simplicity and when schema flexibility is important
  • Avro: Best for systems with evolving schemas and when compatibility is critical
  • Protocol Buffers: Best for performance-critical systems with structured data
  • Lambda Default: Best for simple string-based messages or when using Lambda's built-in deserialization

Advanced

Accessing record metadata

Each Kafka record contains important metadata that you can access alongside the deserialized message content. This metadata helps with message processing, troubleshooting, and implementing advanced patterns like exactly-once processing.

=== "Working with Record Metadata"

```java
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.common.header.Header;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;

public class MetadataKafkaHandler implements RequestHandler<ConsumerRecords<String, Customer>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(MetadataKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_AVRO)
    public String handleRequest(ConsumerRecords<String, Customer> records, Context context) {
        for (ConsumerRecord<String, Customer> record : records) {
            // Log record coordinates for tracing
            LOGGER.info("Processing message from topic '{}'", record.topic());
            LOGGER.info("  Partition: {}, Offset: {}", record.partition(), record.offset());
            LOGGER.info("  Produced at: {}", record.timestamp());

            // Process message headers
            if (record.headers() != null) {
                for (Header header : record.headers()) {
                    LOGGER.info("  Header: {} = {}",
                        header.key(), new String(header.value()));
                }
            }

            // Access the Avro deserialized message content
            Customer customer = record.value(); // Customer class is auto-generated from Avro schema
            LOGGER.info("Processing order for: {}", customer.getName());
            LOGGER.info("Order total: ${}", customer.getOrderTotal());
        }
        return "OK";
    }
}
```

Available metadata properties

Property Description Example Use Case
topic() Topic name the record was published to Routing logic in multi-topic consumers
partition() Kafka partition number Tracking message distribution
offset() Position in the partition De-duplication, exactly-once processing
timestamp() Unix timestamp when record was created Event timing analysis
timestampType() Timestamp type (CREATE_TIME or LOG_APPEND_TIME) Data lineage verification
headers() Key-value pairs attached to the message Cross-cutting concerns like correlation IDs
key() Deserialized message key Customer ID or entity identifier
value() Deserialized message content The actual business data

Error handling

Handle errors gracefully when processing Kafka messages to ensure your application maintains resilience and provides clear diagnostic information. The Kafka utility integrates with standard Java exception handling patterns.

!!! info "Treating Deserialization errors" Read Deserialization failures. Deserialization failures will fail the whole batch and do not execute your handler.

=== "Error Handling"

```java
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.metrics.FlushMetrics;
import software.amazon.lambda.powertools.metrics.Metrics;
import software.amazon.lambda.powertools.metrics.MetricsFactory;
import software.amazon.lambda.powertools.metrics.model.MetricUnit;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class ErrorHandlingKafkaHandler implements RequestHandler<ConsumerRecords<String, Order>, String> {

    private static final Logger LOGGER = LoggerFactory.getLogger(ErrorHandlingKafkaHandler.class);
    private static final Metrics metrics = MetricsFactory.getMetricsInstance();

    @Override
    @Logging
    @FlushMetrics(namespace = "KafkaProcessing", service = "order-processing")
    @Deserialization(type = DeserializationType.KAFKA_AVRO)
    public String handleRequest(ConsumerRecords<String, Order> records, Context context) {
        metrics.addMetric("TotalRecords", records.count(), MetricUnit.COUNT);
        int successfulRecords = 0;
        int failedRecords = 0;

        for (ConsumerRecord<String, Order> record : records) {
            try {
                Order order = record.value(); // Order class is auto-generated from Avro schema
                processOrder(order);
                successfulRecords++;
                metrics.addMetric("ProcessedRecords", 1, MetricUnit.COUNT);
            } catch (Exception e) {
                failedRecords++;
                LOGGER.error("Error processing Kafka message from topic: {}, partition: {}, offset: {}",
                    record.topic(), record.partition(), record.offset(), e);
                metrics.addMetric("ProcessingErrors", 1, MetricUnit.COUNT);
                // Optionally send to DLQ or error topic
                sendToDlq(record);
            }
        }

        return String.format("Processed %d records successfully, %d failed",
            successfulRecords, failedRecords);
    }

    private void processOrder(Order order) {
        // Your business logic here
        LOGGER.info("Processing order: {}", order.getOrderId());
    }

    private void sendToDlq(ConsumerRecord<String, Order> record) {
        // Implementation to send failed records to dead letter queue
    }
}
```

Integrating with Idempotency

When processing Kafka messages in Lambda, failed batches can result in message reprocessing. The idempotency utility prevents duplicate processing by tracking which messages have already been handled, ensuring each message is processed exactly once.

The Idempotency utility automatically stores the result of each successful operation, returning the cached result if the same message is processed again, which prevents potentially harmful duplicate operations like double-charging customers or double-counting metrics.

=== "Idempotent Kafka Processing"

```java
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.idempotency.Idempotency;
import software.amazon.lambda.powertools.idempotency.IdempotencyConfig;
import software.amazon.lambda.powertools.idempotency.Idempotent;
import software.amazon.lambda.powertools.idempotency.persistence.dynamodb.DynamoDBPersistenceStore;
import software.amazon.lambda.powertools.logging.Logging;

public class IdempotentKafkaHandler implements RequestHandler<ConsumerRecords<String, Payment>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(IdempotentKafkaHandler.class);

    public IdempotentKafkaHandler() {
        // Configure idempotency with DynamoDB persistence store
        Idempotency.config()
            .withPersistenceStore(
                DynamoDBPersistenceStore.builder()
                    .withTableName("IdempotencyTable")
                    .build())
            .configure();
    }

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<String, Payment> records, Context context) {
        for (ConsumerRecord<String, Payment> record : records) {
            // Payment class deserialized from JSON
            Payment payment = record.value();

            // Process each message with idempotency protection
            processPayment(payment);
        }
        return "OK";
    }

    @Idempotent
    private void processPayment(Payment payment) {
        LOGGER.info("Processing payment {}", payment.getPaymentId());

        // Your business logic here
        PaymentService.process(payment.getPaymentId(), payment.getCustomerId(), payment.getAmount());
    }
}
```

???+ tip "Ensuring exactly-once processing" The @Idempotent annotation will use the JSON representation of the Payment object to make sure that the same object is only processed exactly once. Even if a batch fails and Lambda retries the messages, each unique payment will be processed exactly once.

Best practices

Batch size configuration

The number of Kafka records processed per Lambda invocation is controlled by your Event Source Mapping configuration. Properly sized batches optimize cost and performance.

=== "Batch size configuration"

```yaml
Resources:
  OrderProcessingFunction:
    Type: AWS::Serverless::Function
    Properties:
      Handler: org.example.OrderHandler::handleRequest
      Runtime: java21
      Events:
        KafkaEvent:
          Type: MSK
          Properties:
            Stream: !GetAtt OrdersMSKCluster.Arn
            Topics:
              - order-events
              - payment-events
            # Configuration for optimal throughput/latency balance
            BatchSize: 100
            MaximumBatchingWindowInSeconds: 5
            StartingPosition: LATEST
            # Enable partial batch success reporting
            FunctionResponseTypes:
              - ReportBatchItemFailures
```

Different workloads benefit from different batch configurations:

  • High-volume, simple processing: Use larger batches (100-500 records) with short timeout
  • Complex processing with database operations: Use smaller batches (10-50 records)
  • Mixed message sizes: Set appropriate batching window (1-5 seconds) to handle variability

Cross-language compatibility

When using binary serialization formats across multiple programming languages, ensure consistent schema handling to prevent deserialization failures.

=== "Using Python naming convention"

```java hl_lines="33 36 39 42 56"
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.RequestHandler;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import software.amazon.lambda.powertools.kafka.Deserialization;
import software.amazon.lambda.powertools.kafka.DeserializationType;
import software.amazon.lambda.powertools.logging.Logging;
import com.fasterxml.jackson.annotation.JsonProperty;
import java.time.Instant;

public class CrossLanguageKafkaHandler implements RequestHandler<ConsumerRecords<String, OrderEvent>, String> {
    private static final Logger LOGGER = LoggerFactory.getLogger(CrossLanguageKafkaHandler.class);

    @Override
    @Logging
    @Deserialization(type = DeserializationType.KAFKA_JSON)
    public String handleRequest(ConsumerRecords<String, OrderEvent> records, Context context) {
        for (ConsumerRecord<String, OrderEvent> record : records) {
            OrderEvent order = record.value(); // OrderEvent class handles JSON with Python field names
            LOGGER.info("Processing order {} from {}",
                order.getOrderId(), order.getOrderDate());
        }
        return "OK";
    }
}

// Example class that handles Python snake_case field names
public class OrderEvent {
    @JsonProperty("order_id")
    private String orderId;

    @JsonProperty("customer_id")
    private String customerId;

    @JsonProperty("total_amount")
    private double totalAmount;

    @JsonProperty("order_date")
    private long orderDateMillis;

    // Getters and setters
    public String getOrderId() { return orderId; }
    public void setOrderId(String orderId) { this.orderId = orderId; }

    public String getCustomerId() { return customerId; }
    public void setCustomerId(String customerId) { this.customerId = customerId; }

    public double getTotalAmount() { return totalAmount; }
    public void setTotalAmount(double totalAmount) { this.totalAmount = totalAmount; }

    public Instant getOrderDate() {
        return Instant.ofEpochMilli(orderDateMillis);
    }
    public void setOrderDate(long orderDateMillis) {
        this.orderDateMillis = orderDateMillis;
    }
}
```

Common cross-language challenges to address:

  • Field naming conventions: camelCase in Java vs snake_case in Python
  • Date/time: representation differences
  • Numeric precision handling: especially decimals

Troubleshooting

Deserialization failures

The Java Kafka utility registers a custom Lambda serializer that performs eager deserialization of all records in the batch before your handler method is invoked.

This means that if any record in the batch fails deserialization, a RuntimeException will be thrown with a concrete error message explaining why deserialization failed, and your handler method will never be called.

Key implications:

  • Batch-level failure: If one record fails deserialization, the entire batch fails
  • Early failure detection: Deserialization errors are caught before your business logic runs
  • Clear error messages: The RuntimeException provides specific details about what went wrong
  • No partial processing: You cannot process some records while skipping failed ones within the same batch

Example of deserialization failure:

// If any record in the batch has invalid Avro data, you'll see:
// RuntimeException: Failed to deserialize Kafka record: Invalid Avro schema for record at offset 12345

!!! warning "Handler method not invoked on deserialization failure" When deserialization fails, your handleRequest method will not be invoked at all. The RuntimeException is thrown before your handler code runs, preventing any processing of the batch.

Handling deserialization failures:

Since deserialization happens before your handler is called, you cannot catch these exceptions within your handler method. Instead, configure your Event Source Mapping with appropriate error handling:

  • Dead Letter Queue (DLQ): Configure a DLQ to capture failed batches for later analysis
  • Maximum Retry Attempts: Set appropriate retry limits to avoid infinite retries
  • Batch Size: Use smaller batch sizes to minimize the impact of individual record failures
# Example SAM template configuration for error handling
Events:
  KafkaEvent:
    Type: MSK
    Properties:
      # ... other properties
      BatchSize: 10 # Smaller batches reduce failure impact
      MaximumRetryAttempts: 3
      DestinationConfig:
        OnFailure:
          Type: SQS
          Destination: !GetAtt DeadLetterQueue.Arn

Schema compatibility issues

Schema compatibility issues often manifest as successful connections but failed deserialization. Common causes include:

  • Schema evolution without backward compatibility: New producer schema is incompatible with consumer schema
  • Field type mismatches: For example, a field changed from String to Integer across systems
  • Missing required fields: Fields required by the consumer schema but absent in the message
  • Default value discrepancies: Different handling of default values between languages

When using Schema Registry, verify schema compatibility rules are properly configured for your topics and that all applications use the same registry.

Memory and timeout optimization

Lambda functions processing Kafka messages may encounter resource constraints, particularly with large batches or complex processing logic.

For memory errors:

  • Increase Lambda memory allocation, which also provides more CPU resources
  • Process fewer records per batch by adjusting the BatchSize parameter in your event source mapping
  • Consider optimizing your message format to reduce memory footprint

For timeout issues:

  • Extend your Lambda function timeout setting to accommodate processing time
  • Implement chunked or asynchronous processing patterns for time-consuming operations
  • Monitor and optimize database operations, external API calls, or other I/O operations in your handler

???+ tip "Monitoring memory usage" Use CloudWatch metrics to track your function's memory utilization. If it consistently exceeds 80% of allocated memory, consider increasing the memory allocation or optimizing your code.

Kafka workflow

Using ESM with Schema Registry validation (SOURCE)

```mermaid sequenceDiagram participant Kafka participant ESM as Event Source Mapping participant SchemaRegistry as Schema Registry participant Lambda participant KafkaUtility participant YourCode Kafka->>+ESM: Send batch of records ESM->>+SchemaRegistry: Validate schema SchemaRegistry-->>-ESM: Confirm schema is valid ESM->>+Lambda: Invoke with validated records (still encoded) Lambda->>+KafkaUtility: Pass Kafka event KafkaUtility->>KafkaUtility: Parse event structure loop For each record KafkaUtility->>KafkaUtility: Decode base64 data KafkaUtility->>KafkaUtility: Deserialize based on DeserializationType end KafkaUtility->>+YourCode: Provide ConsumerRecords YourCode->>YourCode: Process records YourCode-->>-KafkaUtility: Return result KafkaUtility-->>-Lambda: Pass result back Lambda-->>-ESM: Return response ESM-->>-Kafka: Acknowledge processed batch ```

Using ESM with Schema Registry deserialization (JSON)

```mermaid sequenceDiagram participant Kafka participant ESM as Event Source Mapping participant SchemaRegistry as Schema Registry participant Lambda participant KafkaUtility participant YourCode Kafka->>+ESM: Send batch of records ESM->>+SchemaRegistry: Validate and deserialize SchemaRegistry->>SchemaRegistry: Deserialize records SchemaRegistry-->>-ESM: Return deserialized data ESM->>+Lambda: Invoke with pre-deserialized JSON records Lambda->>+KafkaUtility: Pass Kafka event KafkaUtility->>KafkaUtility: Parse event structure loop For each record KafkaUtility->>KafkaUtility: Decode base64 data KafkaUtility->>KafkaUtility: Record is already deserialized KafkaUtility->>KafkaUtility: Map to POJO (if specified) end KafkaUtility->>+YourCode: Provide ConsumerRecords YourCode->>YourCode: Process records YourCode-->>-KafkaUtility: Return result KafkaUtility-->>-Lambda: Pass result back Lambda-->>-ESM: Return response ESM-->>-Kafka: Acknowledge processed batch ```

Using ESM without Schema Registry integration

```mermaid sequenceDiagram participant Kafka participant Lambda participant KafkaUtility participant YourCode Kafka->>+Lambda: Invoke with batch of records (direct integration) Lambda->>+KafkaUtility: Pass raw Kafka event KafkaUtility->>KafkaUtility: Parse event structure loop For each record KafkaUtility->>KafkaUtility: Decode base64 data KafkaUtility->>KafkaUtility: Deserialize based on DeserializationType end KafkaUtility->>+YourCode: Provide ConsumerRecords YourCode->>YourCode: Process records YourCode-->>-KafkaUtility: Return result KafkaUtility-->>-Lambda: Pass result back Lambda-->>-Kafka: Acknowledge processed batch ```

Testing your code

Testing Kafka consumer functions is straightforward with JUnit. You can construct Kafka ConsumerRecords in the default way provided by the kafka-clients library without needing a real Kafka cluster.

=== "Testing your code"

```java
package org.example;

import com.amazonaws.services.lambda.runtime.Context;
import com.amazonaws.services.lambda.runtime.events.KafkaEvent;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.common.TopicPartition;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.extension.ExtendWith;
import org.mockito.Mock;
import org.mockito.junit.jupiter.MockitoExtension;
import java.util.*;

import static org.junit.jupiter.api.Assertions.*;
import static org.mockito.Mockito.*;

@ExtendWith(MockitoExtension.class)
class KafkaHandlerTest {

    @Mock
    private Context context;

    @Test
    void testProcessJsonMessage() {
        // Create a test Kafka event with JSON data
        Order testOrder = new Order("12345", 99.95);
        ConsumerRecord<String, Order> record = new ConsumerRecord<>(
            "orders-topic", 0, 15L, null, testOrder);

        Map<TopicPartition, List<ConsumerRecord<String, Order>>> recordsMap = new HashMap<>();
        recordsMap.put(new TopicPartition("orders-topic", 0), Arrays.asList(record));
        ConsumerRecords<String, Order> records = new ConsumerRecords<>(recordsMap);

        // Create handler and invoke
        JsonKafkaHandler handler = new JsonKafkaHandler();
        String response = handler.handleRequest(records, context);

        // Verify the response
        assertEquals("OK", response);
    }

    @Test
    void testProcessMultipleRecords() {
        // Create a test event with multiple records
        Customer customer1 = new Customer("A1", "Alice");
        Customer customer2 = new Customer("B2", "Bob");

        List<ConsumerRecord<String, Customer>> recordList = Arrays.asList(
            new ConsumerRecord<>("customers-topic", 0, 10L, null, customer1),
            new ConsumerRecord<>("customers-topic", 0, 11L, null, customer2)
        );

        Map<TopicPartition, List<ConsumerRecord<String, Customer>>> recordsMap = new HashMap<>();
        recordsMap.put(new TopicPartition("customers-topic", 0), recordList);
        ConsumerRecords<String, Customer> records = new ConsumerRecords<>(recordsMap);

        // Create handler and invoke
        JsonKafkaHandler handler = new JsonKafkaHandler();
        String response = handler.handleRequest(records, context);

        // Verify the response
        assertEquals("OK", response);
    }
}
```

Extra Resources

Lambda Custom Serializers Compatibility

This Kafka utility uses Lambda custom serializers to provide automatic deserialization of Kafka messages.

Important compatibility considerations:

  • Existing custom serializers: This utility will not be compatible if you already use your own custom Lambda serializer in your project
  • Non-Kafka handlers: Installing this library will not affect default Lambda serialization behavior for non-Kafka related handlers
  • Kafka-specific: The custom serialization only applies to handlers annotated with @Deserialization
  • Lambda default fallback: Using @Deserialization(type = DeserializationType.LAMBDA_DEFAULT) will proxy to Lambda's default serialization behavior

Need help with compatibility?

If you are blocked from adopting this utility due to existing custom serializers or other compatibility concerns, please contact us with your specific use-cases. We'd like to understand your requirements and explore potential solutions.

For more information about Lambda custom serialization, see the official AWS documentation.