A schema ID is a unique identifier assigned when a schema is registered with Schema Registry. Kafka messages carry the ID with serialized data rather than carrying the complete schema, reducing duplication while giving consumers a consistent lookup mechanism. section 6.4.1 section 6.1
The producer validates data against a schema, serializes it, and embeds the schema ID in the message. The consumer extracts the ID, retrieves the corresponding schema from Schema Registry, or uses a cached copy, and deserializes the data. Brokers treat messages as raw bytes and do not validate either the ID or the payload. section 6.4.1 section 6.1
Looking up an identifier
A registry request such as `curl http://schema-registry:8081/schemas/ids/1` retrieves the schema associated with ID 1. A consumer can then use that schema to interpret the serialized message. section 6.4.1
Schema IDs make centralized schema management consistent across producers and consumers, but they are not self-describing business metadata: consumers still need prior knowledge of the message structure. IDs can also differ between development and production, complicating cross-cluster migration and sometimes requiring byte-level identifier replacement. section 6.1 section 6.4.6
