Data Engineering Real-Time Streaming Architectures 2

Question 1

In Apache Kafka, what guarantees message ordering?

Accepted Answer

Ordering within a single partition

Answer

Kafka only guarantees order within a partition, not across an entire topic.

Question 2

What is the primary purpose of a watermark in stream processing?

Accepted Answer

To track progress of event time and handle late data

Answer

Watermarks signal how far event time has advanced so windows can close and handle lateness.

Question 3

Which delivery semantic ensures a message is processed once and only once, even after failures?

Accepted Answer

Exactly-once

Answer

Exactly-once semantics guarantee each record affects state precisely one time despite failures.

Question 4

A tumbling window in stream processing is characterized by:

Accepted Answer

Fixed-size, non-overlapping intervals

Answer

Tumbling windows partition time into fixed, contiguous, non-overlapping buckets.

Question 5

What does backpressure refer to in a streaming pipeline?

Accepted Answer

A downstream consumer signaling it cannot keep up with input rate

Answer

Backpressure is the mechanism by which slow consumers throttle faster producers.

Data Engineering Practice Test