Kappa architecture (streaming only) in Hadoop - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Core principle of Kappa architecture

Which statement best describes the core principle of the Kappa architecture in streaming data processing?

A It processes data only once through a single stream processing pipeline without separate batch layers.

B It requires separate batch and speed layers to handle data processing.

C It stores all data in a data lake before processing in batches.

D It uses multiple batch jobs to update the data warehouse periodically.

💡 Hint

Think about how Kappa architecture simplifies data processing by avoiding multiple layers.

Kappa architecture processes data only once through a single streaming pipeline, eliminating the need for separate batch layers.

❓ Predict Output

intermediate

Output of a streaming filter in Kappa architecture

Given the following Hadoop streaming code snippet that filters events with value > 50, what is the output for input events [30, 60, 45, 80]?

Hadoop

input_events = [30, 60, 45, 80]
filtered_events = list(filter(lambda x: x > 50, input_events))
print(filtered_events)

A [60, 80]

B [30, 45]

C [30, 60, 45, 80]

D []

💡 Hint

Filter keeps only values greater than 50.

The filter function keeps only values greater than 50, so 60 and 80 remain.

❓ data_output

advanced

Result of windowed aggregation in streaming

In a Kappa streaming pipeline, a 5-minute tumbling window sums event values arriving every minute: [2, 3, 5, 7, 1]. What is the sum output for the window?

A 7

B 5

C 10

D 18

💡 Hint

Sum all values in the 5-minute window.

Sum of 2 + 3 + 5 + 7 + 1 equals 18.

🔧 Debug

advanced

Identify error in streaming data processing code

What error will this Hadoop streaming Python code raise?

events = [10, 20, 30]
result = sum(events.filter(lambda x: x > 15))
print(result)

A SyntaxError: invalid syntax

B TypeError: 'int' object is not callable

C AttributeError: 'list' object has no attribute 'filter'

D No error, output is 50

💡 Hint

Check if 'filter' is a method of list objects in Python.

Lists in Python do not have a 'filter' method; 'filter' is a built-in function.

🚀 Application

expert

Choosing Kappa architecture for a use case

You have a system that requires real-time fraud detection on transaction data with minimal latency and no need for batch reprocessing. Which architecture is best suited?

A Data lake architecture, because it stores raw data for later analysis.

B Kappa architecture, because it processes data only once in a streaming pipeline.

C Batch processing architecture, because it handles large volumes efficiently.

D Lambda architecture, because it combines batch and speed layers for accuracy.

💡 Hint

Consider the need for low latency and no batch reprocessing.

Kappa architecture suits real-time needs with a single streaming pipeline, avoiding batch layers.