Partition concept in Kafka - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of this Kafka partition assignment code?

Consider a Kafka consumer group with 3 consumers and a topic with 6 partitions. The partitions are assigned evenly. What partitions will consumer 2 receive?

Kafka

consumers = ['consumer1', 'consumer2', 'consumer3']
partitions = list(range(6))
assignments = {c: [] for c in consumers}
for i, p in enumerate(partitions):
    assignments[consumers[i % len(consumers)]].append(p)
print(assignments['consumer2'])

A [0, 3]

B [2, 5]

C [1, 3, 5]

D [1, 4]

💡 Hint

Partitions are assigned in a round-robin fashion based on the consumer index.

Partitions 0, 3 go to consumer1; 1, 4 go to consumer2; 2, 5 go to consumer3. So consumer2 gets partitions 1 and 4.

🧠 Conceptual

intermediate

How does Kafka decide which partition a message goes to when a key is provided?

In Kafka, when producing a message with a key, how is the partition chosen?

A Kafka hashes the key and assigns the message to the partition based on the hash modulo the number of partitions.

B Kafka always sends messages with keys to partition 0.

C Kafka randomly assigns the message to any partition regardless of the key.

D Kafka assigns messages with keys to partitions in a round-robin manner.

💡 Hint

Think about how consistent partitioning is achieved for messages with the same key.

Kafka uses a hash function on the key and then takes modulo with the number of partitions to ensure messages with the same key go to the same partition.

❓ Predict Output

advanced

What error does this Kafka producer code raise?

This code tries to send a message to a non-existent partition. What error will it raise?

Kafka

from kafka import KafkaProducer
producer = KafkaProducer(bootstrap_servers='localhost:9092')
producer.send('my_topic', key=b'key1', value=b'value1', partition=10)
producer.flush()

A kafka.errors.UnknownTopicOrPartitionError

B kafka.errors.KafkaTimeoutError

C ValueError: Partition index out of range

D No error, message sent successfully

💡 Hint

Check what happens if you specify a partition index that does not exist in the topic.

Kafka raises UnknownTopicOrPartitionError when the specified partition does not exist for the topic.

🚀 Application

advanced

How many partitions will be assigned to each consumer in this group?

A Kafka topic has 7 partitions. There are 4 consumers in the same group consuming this topic. How many partitions will each consumer get?

A Consumers 1 to 3 get 1 partition each, consumer 4 gets 4 partitions

B Consumers 1 to 3 get 2 partitions each, consumer 4 gets 1 partition

C Each consumer gets 2 partitions, and 1 partition is assigned to a random consumer

D Each consumer gets exactly 1 partition, and 3 partitions remain unassigned

💡 Hint

Partitions are distributed as evenly as possible among consumers.

7 partitions divided among 4 consumers means 3 consumers get 2 partitions, and 1 consumer gets 1 partition.

🧠 Conceptual

expert

What is the impact of increasing the number of partitions on message ordering in Kafka?

If you increase the number of partitions for a Kafka topic, what happens to the ordering guarantees of messages with the same key?

A Ordering is preserved globally across all partitions regardless of key.

B Ordering is lost because messages with the same key can be sent to different partitions after increasing partitions.

C Ordering is preserved per key because messages with the same key always go to the same partition.

D Ordering is preserved only if the number of partitions is a power of two.

💡 Hint

Think about how Kafka uses keys to assign partitions and maintain order.

Kafka guarantees ordering per partition. Since keys map to partitions via hashing, messages with the same key stay in the same partition even if partitions increase.