You cannot directly attach a running EC2 instance to a newly created Auto Scaling Group in a production-safe way.
The recommended approach is:
Create an AMI from the existing instance
Create a Launch Template using that AMI
Create a new Auto Scaling Group using the Launch Template
This ensures scalable, consistent, and reproducible infrastructure.
Auto Scaling Groups (ASG) are designed to launch instances using:
Launch Templates
Launch Configurations
They are not designed to manage manually created standalone EC2 instances directly in a scalable manner.
Although AWS allows attaching an existing instance to an ASG using the console or CLI, this is not recommended in production, because:
Instance configuration may not match Launch Template
Future scaling launches may differ
Configuration drift issues may occur
No guaranteed repeatability
Go to EC2 → Select Instance → Actions → Image → Create Image
This captures:
OS
Installed applications
Configuration
Patches
Software stack
Go to EC2 → Launch Templates → Create Launch Template
Specify:
AMI created above
Instance type
Security Groups
IAM Role
Key pair
User Data (if needed)
Go to Auto Scaling → Create ASG
Attach:
Launch Template
Desired Capacity
Min/Max capacity
Target Group (if behind ALB)
Scaling policies
You can attach an existing instance to ASG:
But this does not make the instance part of the Launch Template configuration.
This method is useful only for:
Temporary migration
Gradual transition
A company has a manually configured EC2 web server running in production.
Now traffic is increasing and they want Auto Scaling.
Verified application configuration
Cleaned unnecessary logs and temporary files
Created AMI from running instance
Created Launch Template
Created Auto Scaling Group with min=2, max=10
Attached ALB target group
Tested scale-out using CPU stress
Traffic spike handled automatically
No downtime
Infrastructure became scalable
Reduced manual intervention
Because:
Not reproducible
Risk of config drift
Future instances may differ
Breaks IaC principle
You should:
Move data to EFS/S3/RDS
Make instance stateless before scaling
✔ Never scale stateful instances
✔ Always use golden AMI
✔ Store session data in Redis/ElastiCache
✔ Keep servers stateless
“In production, we never directly attach a running EC2 instance to Auto Scaling. Instead, we create an AMI, build a Launch Template, and deploy an Auto Scaling Group to ensure consistency, scalability, and repeatable infrastructure.”
Not a member yet? Register now
Are you a member? Login now