Why stream processing transforms data in Kafka - Challenge Your Understanding

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of this Kafka Streams transformation?

Consider a Kafka Streams application that reads a stream of user clicks and transforms the data to count clicks per user.

streamsBuilder.stream("clicks")
  .map((key, value) -> KeyValue.pair(value.getUserId(), 1))
  .groupByKey()
  .count()
  .toStream()
  .foreach((userId, count) -> System.out.println(userId + ": " + count));

If the input stream has these records:

• key=null, value={userId: "alice"}
• key=null, value={userId: "bob"}
• key=null, value={userId: "alice"}

What will be printed?

Kafka

streamsBuilder.stream("clicks")
  .map((key, value) -> KeyValue.pair(value.getUserId(), 1))
  .groupByKey()
  .count()
  .toStream()
  .foreach((userId, count) -> System.out.println(userId + ": " + count));

A

alice: 2
bob: 1

B null: 3

C

alice: 1
bob: 1

D No output, because key is null

💡 Hint

Think about how the map step changes the key to the userId, enabling grouping by user.

The map step changes each record's key to the userId, so grouping by key counts clicks per user. Alice appears twice, so her count is 2; Bob appears once, so his count is 1.

🧠 Conceptual

intermediate

Why do stream processing systems transform data?

Why is it important for stream processing systems like Kafka Streams to transform data as it flows through the pipeline?

A To store all raw data permanently without changes

B To convert data into images for visualization

C To enrich, filter, or aggregate data in real-time for faster insights

D To slow down the data flow for batch processing later

💡 Hint

Think about what real-time data processing aims to achieve.

Stream processing transforms data to enrich, filter, or aggregate it immediately, enabling quick decisions and insights without waiting for batch jobs.

🔧 Debug

advanced

Identify the error in this Kafka Streams transformation code

Look at this Kafka Streams code snippet:

streamsBuilder.stream("orders")
  .filter((key, value) -> value.getAmount() > 100)
  .mapValues(value -> value.getCustomerName())
  .to("high-value-customers");

What error will this code cause when run?

A Compilation error due to missing semicolon

B Logical error: filter condition is reversed

C No error, code runs correctly

D Runtime error because <code>to</code> requires a key serializer but key is null

💡 Hint

Consider what happens to the key after mapValues and what to expects.

The mapValues changes only the value, but the key remains the same. If the original key is null or not serialized properly, to will cause a runtime serialization error.

📝 Syntax

advanced

Which option correctly transforms a stream to uppercase values?

Given a Kafka Streams input stream of strings, which code snippet correctly transforms all values to uppercase?

A streamsBuilder.stream("input").mapValues(value -> value.toUpperCase()).to("output");

B streamsBuilder.stream("input").map((key, value) -> value.toUpperCase()).to("output");

C streamsBuilder.stream("input").mapValues(value -> value.uppercase()).to("output");

D streamsBuilder.stream("input").mapValues(value -> value.toUpper()).to("output");

💡 Hint

Remember the correct Java method name for uppercase conversion.

The correct method is toUpperCase(). Option A uses mapValues which transforms only the value, keeping the key unchanged.

🚀 Application

expert

How does stream processing enable real-time fraud detection?

Imagine a bank uses Kafka Streams to detect fraudulent transactions in real-time. Which transformation best supports this use case?

A Batch processing transactions once a day for fraud reports

B Filtering transactions above a threshold and aggregating counts per account within a time window

C Storing all transactions in a database without processing

D Converting transactions to JSON strings without filtering

💡 Hint

Think about how to spot suspicious activity quickly using streaming data.

Filtering high-value transactions and aggregating counts per account in short time windows helps detect unusual patterns immediately, enabling fast fraud alerts.