Compute - EC2 Auto Scaling

Question 1

What is EC2 Auto Scaling?

Answer 1

• EC2 Auto Scaling is a mechanism that automatically adjusts the number of EC2 instances in a fleet based on custom-defined metrics and thresholds to meet the demand.

Question 2

What services does the Auto Scaling service support?

Answer 2

• The Auto Scaling service supports scaling of Amazon ECS tasks, DynamoDB tables and indexes, and Amazon Aurora replicas, in addition to EC2 Auto Scaling. However, this course focuses only on EC2 Auto Scaling.

Question 3

What is an example use case for EC2 Auto Scaling?

Answer 3

• An example use case is when you have a single EC2 instance acting as a web server and the demand increases, causing increased CPU utilization. With Auto Scaling, you can configure a metric to automatically launch a second instance when the CPU utilization reaches a certain threshold, effectively load balancing the traffic and improving performance.

Question 4

How does Auto Scaling help optimize the cost of EC2 resources?

Answer 4

• By automatically scaling the resources based on demand, Auto Scaling allows you to increase resources when needed and decrease them when the demand decreases. This helps optimize costs as you only pay for resources when they are running.

Question 5

What are some advantages of using EC2 Auto Scaling?

Answer 5

• Automation: Automatic provisioning based on custom-defined thresholds eliminates the need for manual resource deployment.
• Greater customer satisfaction: Provisioning enough capacity to meet increased demand ensures good performance and user retention.
• Cost reduction: Scaling down resources when demand decreases saves costs by stopping payments for unnecessary resources.
• Scalable and flexible architecture: When combined with an Elastic Load Balancer, Auto Scaling enables building a scalable and flexible infrastructure.

Question 6

How does Auto Scaling work with an Elastic Load Balancer?

Answer 6

• When coupled with an Elastic Load Balancer, Auto Scaling distributes incoming traffic evenly across multiple instances, improving performance and reliability by load balancing the demand.