Schema Registry concept in Kafka - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of this Schema Registry API call?

Given a Schema Registry REST API call to get the latest schema for a subject, what is the expected output?

Kafka

curl -X GET http://localhost:8081/subjects/user-value/versions/latest

A {"subjects":["user-key","user-value"]}

B {"error_code":404,"message":"Subject not found"}

C {"id":10,"schema":"{\"type\":\"string\"}"}

D {"subject":"user-value","version":3,"id":10,"schema":"{\"type\":\"record\",\"name\":\"User\",\"fields\":[{\"name\":\"name\",\"type\":\"string\"}]}"}

💡 Hint

The latest schema for a subject returns detailed schema info including subject, version, id, and schema string.

The GET request to /subjects/{subject}/versions/latest returns the latest schema version details for that subject, including subject name, version number, schema id, and the schema itself as a JSON string.

🧠 Conceptual

intermediate

What is the main purpose of a Schema Registry in Kafka?

Choose the best description of what a Schema Registry does in a Kafka ecosystem.

A It stores and manages schemas for Kafka messages to ensure compatibility and validation.

B It acts as a message broker to route Kafka messages between producers and consumers.

C It provides a user interface to monitor Kafka cluster health and metrics.

D It compresses Kafka messages to reduce network bandwidth usage.

💡 Hint

Think about how Kafka producers and consumers agree on message formats.

Schema Registry stores schemas centrally so producers and consumers can validate and evolve message formats safely, preventing incompatible changes.

🔧 Debug

advanced

Why does this Kafka producer fail when using Schema Registry?

A Kafka producer configured with Schema Registry throws an error when sending messages. What is the likely cause?

Kafka

Producer config:
{
  "bootstrap.servers": "localhost:9092",
  "key.serializer": "org.apache.kafka.common.serialization.StringSerializer",
  "value.serializer": "io.confluent.kafka.serializers.KafkaAvroSerializer",
  "schema.registry.url": "http://localhost:8081"
}

Error message: "org.apache.kafka.common.errors.SerializationException: Error serializing Avro message"

A The key.serializer is incompatible with the value.serializer.

B The Kafka broker is down and cannot accept messages.

C The schema for the message value is not registered in the Schema Registry.

D The producer is missing the bootstrap.servers configuration.

💡 Hint

SerializationException often means the schema is missing or incompatible.

KafkaAvroSerializer requires the schema to be registered in Schema Registry before serializing messages. If the schema is missing, serialization fails.

📝 Syntax

advanced

Which Avro schema is valid for registering in Schema Registry?

Select the valid Avro schema JSON string that can be registered in Schema Registry.

A {"type":"record","name":"User","fields":[{"name":"id","type":"int"},{"name":"name","type":"string"}]}

B {"type":"record","fields":[{"name":"id","type":"int"},{"name":"name","type":"string"}]}

C {"type":"record","name":"User","fields":{"id":"int","name":"string"}}

D {"type":"record","name":"User","fields":[{"id":"int"},{"name":"string"}]}

💡 Hint

Avro record fields must be an array of objects with 'name' and 'type' keys.

Option A correctly defines a record with a name and an array of fields, each with 'name' and 'type'. Other options have missing or wrong structure.

🚀 Application

expert

How many schema versions exist after these operations?

You register a schema for subject 'order-value' three times with these schemas: 1) {"type":"record","name":"Order","fields":[{"name":"id","type":"int"}]} 2) {"type":"record","name":"Order","fields":[{"name":"id","type":"int"},{"name":"amount","type":"float"}]} 3) {"type":"record","name":"Order","fields":[{"name":"id","type":"int"},{"name":"amount","type":"float"}]} How many versions will the subject 'order-value' have in Schema Registry?

A 0

B 2

C 1

D 3

💡 Hint

Schema Registry does not create a new version if the schema is identical to the latest one.

The first registration creates version 1, the second a new version 2 because the schema changed, the third is identical to version 2 so no new version is created.