AWS 9.4 Reliability

Question 1

A system runs on a single large EC2 instance and crashes during peak demand. Which reliability principles are violated and how should the architecture be improved?

Answer 1

Violated:
• Scale horizontally
• Avoid single point of failure

Fix:
• Replace with multiple smaller instances
• Use load balancing + Auto Scaling

Question 2

A company has backup systems but has never tested them. During failure, recovery does not work. Which principle is missing and why is it critical?

Answer 2

Test recovery procedures

It is critical because untested recovery plans often fail in real scenarios.

Question 3

A system detects failure but requires manual intervention to recover. What reliability principle is missing and how should it be implemented?

Answer 3

Automatically recover from failure

Implement using:
• Monitoring + alarms
• Auto recovery mechanisms
• Self-healing systems

Question 4

A company provisions fixed capacity based on estimates and frequently over/under-provisions resources. What principle fixes this and how?

Answer 4

Stop guessing capacity

Use:
• Monitoring metrics
• Auto Scaling to dynamically adjust resources

Question 5

A system handles failures well but breaks whenever updates are deployed. Which reliability concept is being ignored?

Answer 5

Manage change in automation

Changes should be:
• Automated
• Controlled
• Tested before deployment