The session provides a comprehensive overview of a Kafka application state recovery service, specifically designed to optimize recovery times for applications.

In an event-sourced system, an application’s current state is typically derived from the complete history of events. This process involves performing a thorough read of a Kafka topic and transporting the events over the network prior to handling a command. However, this process can become time-consuming, particularly for applications with a high number of messages in the topic. We’ve developed a sophisticated Kafka application state recovery service that efficiently retrieves the state of a topic at any desired time or offset.

In this session, we will explore specific use cases where log compaction-based state recovery proves inadequate and demonstrate how the Kafka application state recovery service addresses these shortcomings with an effective, user-friendly solution.