You want to design a chaos experiment that tests system recovery when 30% of microservice instances fail simultaneously. Which approach best achieves this?

hard📝 Trade-off Q8 of 15

Microservices - Testing Microservices

ARestart all instances one by one without failure

BGradually add 30% more instances to the cluster

CRandomly kill 30% of instances at once and monitor system behavior

DIncrease network bandwidth by 30%

Step-by-Step Solution

Solution:

Step 1: Define failure simulation for 30% instances
To test recovery, kill 30% instances simultaneously to simulate failure.
Step 2: Evaluate options for simulating failure
Only Randomly kill 30% of instances at once and monitor system behavior kills 30% instances at once, triggering recovery mechanisms.
Final Answer:
Randomly kill 30% of instances at once and monitor system behavior -> Option C
Quick Check:
Simulate failure by killing instances [OK]

Quick Trick: Kill target percentage of instances simultaneously [OK]

Common Mistakes:

Confusing failure with scaling
Restarting without failure does not test recovery
Changing bandwidth unrelated to failure

Master "Testing Microservices" in Microservices

9 interactive learning modes - each teaches the same concept differently

Learn Why Deep Arch Try Challenge Design Recall Scale

Want More Practice?

15+ quiz questions · All difficulty levels · Free

Free Signup - Practice All Questions

More Microservices Quizzes

You want to design a chaos experiment that tests system recovery when 30% of microservice instances fail simultaneously. Which approach best achieves this?

Step 1: Define failure simulation for 30% instances

Step 2: Evaluate options for simulating failure

Final Answer:

Quick Check:

Want More Practice?