How do you handle message schema evolution in Kafka?

Schema evolution strategies: 1) Backward compatibility: new schema can read data written with old schema (add fields with defaults, never remove required fields). 2) Forward compatibility: old schema can read data written with new schema. 3) Full compatibility: both directions simultaneously. Use Confluent Schema Registry to enforce compatibility rules and store schemas centrally. Avro and Protobuf handle evolution best. Strategies: add optional fields only, use wrapper types, version topic names for breaking changes.

Expected key concepts: schema evolution, backward compatibility, forward compatibility, Schema Registry, Avro, Protobuf, optional fields, wrapper types, defaults, version topics