Hey everyone. I've enjoyed having Cloudfront taking the heavy load off my public facing servers (a few of them EC2),
But now, the time has come to have scalable application servers, and I want them to spawn based upon need and route traffic through a load balancer.
Here's what I suppose sofar:
1) Your one load balancer sits online the whole time, at 720 hours a month at approximately $20/month
2) The LB receives incoming traffic, and sends it to your EC2 instances.
Goal: How to make the load balancer say "oh, our current EC2 instances have full CPU, I'd better launch another EC2 instance and send traffic to it"
Q1) Can the LB actually bring new EC2 instances online as needed based upon current EC2's being used to capacity?
Q2) Can the LB then spin those new instances down after the load has passed so as to not burn through minutes of uptime?