Auto Scaling

Automatically add capacity under load and remove it when quiet to stay fast and cost-efficient.

Auto scaling adjusts how much compute you run based on demand. Under a traffic spike it adds tasks or instances; when things quiet down it removes them, so you pay only for what you need.

What can scale

  • ECS service auto scaling — add/remove Fargate tasks on CPU or request count.
  • EC2 Auto Scaling Groups — add/remove instances behind a load balancer.
  • Lambda — scales concurrency automatically with no configuration.

Target tracking

The simplest policy is target tracking: pick a metric and a target (e.g. keep average CPU at 50%) and AWS adds or removes capacity to hold it there.

Example

Example · bash
# Keep ECS service CPU near 50% by scaling tasks 2..10
aws application-autoscaling put-scaling-policy \
  --service-namespace ecs \
  --resource-id service/myapp/web \
  --scalable-dimension ecs:service:DesiredCount \
  --policy-name cpu-target \
  --policy-type TargetTrackingScaling \
  --target-tracking-scaling-policy-configuration \
    'TargetValue=50.0,PredefinedMetricSpecification={PredefinedMetricType=ECSServiceAverageCPUUtilization}'

When to use it

  • An e-commerce site uses EC2 Auto Scaling to add instances when CPU exceeds 70% during Black Friday traffic and removes them when traffic subsides.
  • An ECS service uses Application Auto Scaling based on ALB request count per target so container count tracks actual load precisely.
  • A data pipeline uses scheduled scaling to pre-warm extra capacity before a nightly batch job and scale down after it completes.

More examples

Create EC2 Auto Scaling group

Creates an auto scaling group with a min of 2 and max of 10 instances using an existing launch template.

Example · bash
aws autoscaling create-auto-scaling-group \
  --auto-scaling-group-name myapp-asg \
  --launch-template LaunchTemplateId=lt-0abc123,Version='$Latest' \
  --min-size 2 \
  --max-size 10 \
  --desired-capacity 2 \
  --vpc-zone-identifier 'subnet-aaa,subnet-bbb' \
  --health-check-type ELB \
  --health-check-grace-period 60

CPU-based scaling policy

Adds a target-tracking policy that automatically adjusts instance count to keep average CPU around 70%.

Example · bash
aws autoscaling put-scaling-policy \
  --auto-scaling-group-name myapp-asg \
  --policy-name cpu-target-tracking \
  --policy-type TargetTrackingScaling \
  --target-tracking-configuration \
    'PredefinedMetricSpecification={PredefinedMetricType=ASGAverageCPUUtilization},TargetValue=70.0'

ECS Application Auto Scaling

Registers the ECS service as a scalable target and adds CPU-based target-tracking so task count scales with demand.

Example · bash
aws application-autoscaling register-scalable-target \
  --service-namespace ecs \
  --resource-id service/prod/myapp \
  --scalable-dimension ecs:service:DesiredCount \
  --min-capacity 2 \
  --max-capacity 20

aws application-autoscaling put-scaling-policy \
  --service-namespace ecs \
  --resource-id service/prod/myapp \
  --scalable-dimension ecs:service:DesiredCount \
  --policy-name ecs-cpu-tracking \
  --policy-type TargetTrackingScaling \
  --target-tracking-scaling-policy-configuration \
    'PredefinedMetricSpecification={PredefinedMetricType=ECSServiceAverageCPUUtilization},TargetValue=60.0'

Related

deploy cloudwatchdeploy cost

Discussion

  • Be the first to comment on this lesson.