Recipes for Apache Flink®

This is a collection of examples of Apache Flink applications in the format of "recipes". Each recipe illustrates how you can solve a specific problem by leveraging one or more of the APIs of Apache Flink.

Each of these recipes is a self-contained module. They can be a starting point for solving your application requirements with Apache Flink.

Requirements

• Maven
• Java 11

Recipes

APIs and languages

• Batch and streaming with the Table and DataStream APIs
• Writing an application in Kotlin™

Connectors and formats

• Deserializing JSON from Kafka
• Reading Apache Kafka headers
• Continuously reading CSV files
• Reading Google Protocol Buffers
• Change data capture

State and application lifecycle

• Exactly-once with Apache Kafka®
• Upgrading Flink (Table API)
• Migrating state away from Kryo

Event routing

• Creating dead letter queues
• Capturing late data
• Splitting Apache Kafka® event streams

Use cases

• Joining and deduplicating data
• Using session windows
• Using CEP to alert when problems persist
• Measuring latency
• An enrichment join that waits for missing data