ASG Scaling Policies

Question 1

What are the main types of ASG scaling policies?

Answer 1

Target Tracking: Adjusts instances to maintain a target value for a specific metric.
Step Scaling: Changes the instance count in predefined steps, based on the size of the metric alarm breach.
Scheduled Scaling: Scales based on known load patterns at specific times.
Predictive Scaling: Uses machine learning to predict and scale for future demands.

Example: Target Tracking adjusts the number of instances to keep the average CPU utilization at 70%.

Question 2

What triggers ASG scaling actions?

Answer 2

Scaling actions are triggered by Amazon CloudWatch alarms based on metrics such as CPU utilization, network usage, or custom metrics.
Each policy type uses different metrics to initiate scaling to ensure optimal performance and cost efficiency.

Additional Information: CloudWatch alarms can be set to trigger scaling actions based on specific thresholds.

Question 3

What is a cooldown period in ASG scaling?

Answer 3

A cooldown period is a temporary halt in scaling activities after a scaling action is executed.
This period allows the system to stabilize by delaying further scaling actions, ensuring that the previous scaling activity has its intended effect without unnecessary cost or performance impacts.

Example Sentence: After scaling up due to increased network traffic, a 300-second cooldown period was applied to prevent rapid and unnecessary scaling changes.