3.5.2. Elastic Load Balancing (ELB)

Elastic Load Balancing (ELB) Distributes incoming network traffic across multiple instances. Increases fault tolerance by rerouting traffic away from unhealthy instances.

Elastic Load Balancing (ELB) is a service that helps manage incoming network traffic to ensure that no single instance becomes overwhelmed with requests. By distributing the traffic across multiple instances, ELB enhances the reliability of your application. If one instance becomes unhealthy or fails, ELB will automatically redirect the traffic to healthy instances, thus maintaining the availability of your service.

Think of ELB like a traffic director at a busy intersection. Just as the director ensures that cars are evenly distributed between lanes to avoid bottlenecks, ELB ensures that user requests are evenly distributed among multiple servers to maintain smooth operation.

Supports different types: Application Load Balancer: Works at HTTP/HTTPS layer (Layer 7), supports advanced routing. Network Load Balancer: Works at TCP layer (Layer 4), supports ultra-low latency. Classic Load Balancer: Legacy option supporting both Layer 4 and 7.

ELB offers various types of load balancers suited for different use cases. The Application Load Balancer operates at the application layer (Layer 7), allowing for more complex routing decisions based on HTTP requests, which is useful for applications that need advanced features like host-based routing. The Network Load Balancer functions at the transport layer (Layer 4) and is designed for extreme performance and low latency, making it ideal for real-time traffic. Finally, the Classic Load Balancer is an older option that supports both layers but lacks some of the features of the newer types.

Imagine a restaurant with different dining areas. An Application Load Balancer is like a host who directs customers to specific tables based on their preferences (such as a quiet area or a place with a view). A Network Load Balancer is akin to a waiter who swiftly serves customers without any unnecessary delay. The Classic Load Balancer is similar to a general seating arrangement where anyone can sit anywhere without much thought.

ELB sends incoming traffic to healthy instances. Auto Scaling changes the number of instances dynamically. Together, they ensure your application is highly available, scalable, and fault-tolerant.

ELB and Auto Scaling work in tandem to provide a robust infrastructure for your applications. While ELB manages the distribution of traffic to ensure that user requests are handled by healthy instances, Auto Scaling adjusts the number of instances based on the traffic. During high demand, Auto Scaling can launch new instances to handle the additional load, and when demand decreases, it can terminate unnecessary instances to save costs, maintaining efficiency.

Consider a concert venue. ELB is like the ushers who direct attendees to open seats, while Auto Scaling is like the management team that decides to add more rows of chairs when a large crowd shows up and remove extra chairs when the crowd thins out. Together, they ensure that everyone has a seat without wasting resources.

--