Every time a customer's instance reaches capacity, it cuts a sev2 ticket to a new grad l4 who jumps on a computer within 15 mins, scrambles to find an SOP, and manually scales up the cluster while praying they didn't mess up. Seamless autoscaling for the customer!
AWS is managing in the backend for you and handling ops and on-call issues. Nothing is serverless in reality, AWS Lambdas run on EC2 instances but let's say a team of engineers wrote a code and provided APIs that can do certain stuff. It ultimately runs on a CPU managed by AWS where AWS engineers can SSH. Some algorithms keep doing optimization, scale the fleet (optimal use of IOPS) etc. It is no different than saying web app as managed service, instead of a website a code is managing CPUs/resources.
Still curious to know.
Dude you work for Amazon. Really?
Was an intern at Amazon
LMAO