Kafka Topics & Partitions Flashcards

What is a Kafka 'segment' in the context of a partition log?

Answer: A file on disk that holds a contiguous portion of a partition's messages

A segment is a physical log file on the broker's disk; each partition's log is split into multiple segment files for efficient cleanup and management.

Which topic-level configuration controls the maximum size a log segment file can reach before a new one is created?

Answer: log.segment.bytes

log.segment.bytes (default 1 GB) determines the maximum size of a single log segment file before Kafka rolls over to a new segment.

What happens to messages in a tombstone record during log compaction?

Answer: The record with a null value (tombstone) signals deletion of that key during compaction

A tombstone is a message with a null value; during compaction, both the tombstone and any older records for that key are eventually deleted.

What does the 'unclean.leader.election.enable' setting control?

Answer: Whether an out-of-sync replica can be elected leader, risking data loss

When set to true, unclean.leader.election.enable allows a replica not in the ISR to become leader, potentially losing messages that were not yet replicated.

How does Kafka ensure ordering guarantees within a topic?

Answer: Ordering is guaranteed only within a single partition, not across partitions

Kafka guarantees message order only within a single partition; messages across different partitions of the same topic have no global ordering.

What is the role of the Kafka controller broker?

Answer: Manages partition leader elections and broker metadata changes cluster-wide

The controller is a special broker elected among the cluster that is responsible for partition leadership elections and propagating metadata updates.