i have kind of a problem with the load balancer health checks during peak periods.
my service contains 10 instances (tasks) sitting on 4 EC2 machines.
in front