You design a Saga with multiple microservices communicating via Kafka. To ensure scalability and fault tolerance, which approach is best?

hard📝 Application Q15 of 15

Kafka - Event-Driven Architecture

AStore Saga state only in a centralized database without event logs

BUse separate Kafka topics per Saga step and implement idempotent consumers

CUse a single Kafka topic for all Saga events without consumer groups

DSend synchronous HTTP calls between services instead of Kafka events

Step-by-Step Solution

Solution:

Step 1: Consider scalability and fault tolerance needs
Separate topics per step allow independent scaling and isolation of events.
Step 2: Ensure idempotency in consumers
Idempotent consumers prevent duplicate processing on retries, improving fault tolerance.
Step 3: Evaluate other options
Single topic without consumer groups limits scalability; centralized DB without events loses Saga benefits; synchronous calls reduce fault tolerance.
Final Answer:
Use separate Kafka topics per Saga step and implement idempotent consumers -> Option B
Quick Check:
Separate topics + idempotent consumers = scalable Saga [OK]

Quick Trick: Separate topics and idempotent consumers scale Saga well [OK]

Common Mistakes:

MISTAKES

Master "Event-Driven Architecture" in Kafka

9 interactive learning modes - each teaches the same concept differently

Want More Practice?

15+ quiz questions · All difficulty levels · Free

More Kafka Quizzes