I have .NET API Microservice (auto scaling based on CPU Usage) using ngnix that build a message from the GET Request and send it to Kinesis on AWS. Also have Health controller with Get request "/Health" that return OK status (200) that will be using only by the ELB to check the health status of one of my nodes (t3.large).
Under heavy load (1 million request in 1 minute). The ELB stop sending requests to my server because failing at health check. The CPU Usage get around 90%. How can I manage to handle those spikes and pass the ELB Health check?
[