Which statement accurately describes the KTable's relation to the changelog stream?

• A. KTable is an abstraction of the changelog stream from a subset of partitions of a topic.
• B. KTable is an abstraction of the changelog stream from all partitions of a topic.
• C. KTable represents a stream of append-only events.
• D. KTable is a direct replacement for a topic's producer.

Answer: (A)

The question tests how a KTable relates to the changelog stream in a Kafka Streams application. A KTable represents the current state of a table, where for every key you have the latest value, and it is backed by a changelog stream that records updates to that table. In a distributed setup, each application instance only processes the partitions of the changelog topic that it’s assigned to. So, from the perspective of a single instance, a KTable is an abstraction built from a subset of partitions of the topic’s changelog stream. Across the whole application, the collection of instances covers all partitions, yielding the full KTable state.

This is different from a KStream, which is a stream of append-only events, and it isn’t a direct replacement for a topic’s producer—producers write to topics, while a KTable is a stateful view rebuilt from the changelog.