AWS EC2 AutoScaling

• Now we would want to increase/decrease number of ec2 instances in auto scaling group depending on some factors.

• So lets try to increase ec2 instances when the load on CPU increases for this

  • When CPU utilization > 80 % for 5 minutes lets increase one ec2 instance
  • When CPU utilization < 40 % for 5 minutes lets decrease one ec2 instance
  • Minimum ec2 instances => 1
  • Max ec2 instances => 5

• Lets do this activity using launch configuration

• Lets look at step scaling which will be easier to understand

• Now lets look at simple scaling

• To satisfy my application demands

  • When CPU utilization > 80 % for 5 minutes lets increase one ec2 instance
  • When CPU utilization < 40 % for 5 minutes lets decrease one ec2 instance

• Now to create artificial load on cpu, lets login into ec2 instance and run the stress tool.

 stress --cpu 8 --io 4 --vm 2 --vm-bytes 128M --timeout 20m -v

• Now once the alarm is reached
• Now lets look at low cpu
• Next Steps:
  • We need to try autoscaling with launch template
  • We need to add load balancer to balance the traffic across multiple ec2 instances