📡 You're offline — showing cached content
New version available!
Quick Access
Tutorials AWS Solutions Architect Auto Scaling — Elasticity

Auto Scaling — Elasticity

4 min read Quiz at the end
Learn how Auto Scaling Groups maintain application availability and cost efficiency using scaling policies, health checks, and launch templates.

Auto Scaling — Automatically Adjusting Capacity

Auto Scaling automatically adds servers when you need more capacity and removes them when you need less. This keeps your application performing well during traffic spikes and saves money during quiet periods.

Teacher Note: Think of a ride-sharing company. At rush hour, they activate more drivers. At 3am, most drivers log off. Auto Scaling does the same for your servers — automatically.

Auto Scaling Group Configuration

SettingPurposeExample
MinimumAlways keep this many instances running2 (never go below 2)
MaximumNever exceed this many instances10 (cost control)
DesiredTarget number right now4 (current running count)

Scaling Policies

PolicyHow it WorksBest For
Target TrackingMaintain a target metric — e.g. keep CPU at 70%Simple, most recommended
Step ScalingAdd 2 instances when CPU 70-80%, add 5 when CPU 80-90%Granular control
ScheduledScale up at 8am every weekday, scale down at 8pmKnown traffic patterns
PredictiveML predicts load and pre-scales before traffic arrivesRecurring patterns, launch ready

Launch Template vs Launch Configuration

  • Launch Template: NEWER — supports versioning, Spot and On-Demand mix, multiple instance types — USE THIS
  • Launch Configuration: OLDER — being deprecated — avoid in new architectures

ASG Health Checks

  • EC2 Health Check: is the instance running? (default)
  • ELB Health Check: is the application responding correctly on the defined path? (recommended for web apps)
  • If instance fails: ASG terminates it and launches a replacement automatically
Exam Tip: Target Tracking is almost always the correct answer for scaling policies. The most common exam pattern: ALB + ASG with Target Tracking on CPU or RequestCountPerTarget. Scale-in protection prevents instances from being terminated during important operations.
Topic Quiz · 2 questions

Test your understanding before moving on

1. A company wants their web application to always have exactly 70% average CPU utilisation across all EC2 instances. Which Auto Scaling policy achieves this automatically?
💡 Target Tracking Scaling automatically adds or removes instances to maintain the specified target metric value — in this case 70% CPU.
2. An Auto Scaling Group has minimum=2, desired=4, maximum=8. Traffic doubles unexpectedly. What happens?
💡 The maximum setting acts as a hard ceiling — ASG scales up to 8 instances (the maximum) and cannot go beyond. This prevents runaway costs.