Kafka

2 days Online / Zoom English Certificate 15 people max Victor Rentea €7500 ~~9375~~

A comprehensive dive into Kafka best practices, covering fundamentals, Java clients, testing, eventual consistency patterns, Kafka Streams, and performance tuning, with real-world examples and insights to help you master messaging systems and improve your architecture.

Introduction

In this comprehensive workshop, we will explore Kafka from the ground up, covering everything from fundamental concepts to advanced performance tuning. We’ll dive into Java clients, KafkaTemplate, @KafkaListener, and error handling mechanisms such as retries and dead-letter queues. Additionally, we’ll look at event schema management, Avro integration, and schema registry.

By the end of this course, you’ll have the skills and knowledge to tackle real-world Kafka use cases, optimize performance, and integrate Kafka seamlessly into your systems.

Agenda

Fundamentals

Producer, Topic, partition, replication, consumer group/instance/offsets/rebalancing
Fault-tolerance on producer, broker, consumer: core config props
Manage topics, produce, and consume messages using Kafka CLI and IntelliJ Plugin

Java Clients

KafkaProducer / KafkaConsumer in in kafka client
KafkaTemplate and sending patterns
@KafkaListener and error handling: log, sync retry, delayed retry, dead-letter
Spring Cloud Stream Function (optional)
Message Tracing with Spring-Kafka Interceptors
Event Schema: versioning, creating an upcaster/downcaster
Intro to Avro / Schema Registry

Testing

Separating business logic tests from Kafka with spy beans
Embedded Kafka
Testing against a standalone or Testcontainer-ed Kafka Broker

Review Code Samples from Audience

Eventual Consistency Patterns

Handling duplicated and out-of-order messages with Inbox Table
Outbox Table
Change Data Capture with Debezium / Kafka Connect

Kafka Streams

Filter, Transform, Explore a record: Word Counter Kata
Enriching data via API call or joining a KTable/GlobalKTable
Branch and Merge streams
Stateful operations: count, reduce, aggregate
Designing maintainable aggregators (+many real-life applications)
Time windowed aggregations: hopping, tumbling, sliding and session; stream-time and grace
Query a KTable state store (CQRS)
Processor API to read/write in state stores, and schedule tasks vs wall-clock-time
Test Driver and controlling time

Tuning Kafka Performance

Key Metrics to observe/alarm
Performance best practices on Partitions, Producer, Consumer and Broker

Victor is a Java Champion, Java Enterprise Architect, and Independent Trainer, delivering intense masterclasses for companies worldwide since 2014. He is passionate about software architecture, clean code, refactoring, and unit testing. He is a regular speaker at top international conferences. He's also the founder of the Bucharest Software Craftsmanship Community with 4000 members.

🎉 Request training for your team →