Netflix uses Amazon's elastic load balancing service to router the traffic to different front-end services. ELBs are set up in such a way that the load is first balanced across the zones and it moves towards the instances (Two-tier balancing scheme). The first tier contains the basic DNS-based round-robin load balancing so when the request lands, it will be first balanced across zones using round-robin. The second tier of ELB is an array of load balancer instances that does round-robin load balancing.

Elastic Load Balancer

The elastic load balancer is the service provided by Amazon to handle load balancing. ELBs distributes the network traffic and incoming application such as container (ECS), EC2 instances and IP addresses, etc. across multiple targets in different Availability Zones. Elastic Load Balancer monitors the health of its registered targets, accepts the incoming traffic, and distributes the traffic only on healthy targets.

It also provides a feature of deletion protection, enabling it will prevent load balancers from deleting accidentally. Let's assume that we have an application that is getting limited traffic and deployed to the EC2 instance.

With the passage of time, this application encounters expected traffic which increases the load. Hence, we have to allot more resources to the application.

Due to too much traffic on the service, the application might end up crashing. Amazon provides a feature of scaling up and scaling down the resources. Load balancer realized the load, hence it automatically scape up the resources which are sufficient to handle that load.

Elastic Load Balancer basically serves the purpose of the single point of contact for the client. This makes the life to the client very easy where the client sends the request to the load balancer and it then distributes them to the targets like EC2, etc in different Availability Zones.