The feature that helps overburdened traffic return to a normal state is load shedding. Load shedding is a technique used to manage incoming requests and optimize resource allocation during periods of high demand. It involves prioritizing and serving critical requests while temporarily rejecting or delaying non-critical requests. This helps to maintain system availability and prevent overload.
In the context of web services, load shedding can be implemented using various tools and services, such as:
-
Elastic Load Balancing (ELB): ELB helps manage incoming requests by optimally routing traffic to all instances in an Auto Scaling group. It can route traffic to healthy instances and, if necessary, to other availability zones until the original instances restore to a healthy state.
-
Auto Scaling: Auto Scaling dynamically scales resources based on demand and can work in conjunction with ELB to distribute incoming traffic across multiple availability zones. This helps to take advantage of the safety and reliability of geographic redundancy.
-
Health Checks: Monitoring the health state of each instance in an Auto Scaling group using EC2 instance status checks and Elastic Load Balancing health checks helps ensure that only healthy instances are serving traffic. Unhealthy instances can be temporarily removed from the load balancers rotation until they are restored to a healthy state.
-
Domain Name System (DNS) Failover: Integrating Amazon Route 53 and DNS failover can further boost application resiliency by routing traffic to another healthy ELB or failing over across AWS regions in the event of an outage.
