Start with what is Elastic Load Balancing?
Elastic Load Balancing automatically distribute incoming application traffic to multiple destinations, such as Amazon EC2 instances, containers, IP address, and lambda functions. It can handle a variety of application traffic load in a single Availability Zone or multiple Availability Zones.
Key points:
The following two types of load balancer:
1. Internet Facing: has public IP
2. Internal Load Balancer: routes traffic within a VPC
There are three categories of Load Balancer:
1. Application Load Balancer
2. Network Load Balancer
3. Classic Load Balance
First Application Load Balancer
Operates at Layer 7.
Supports WebSockets and Secure WebSockets.
Supports SNI.
Supports IPv6.
Second Network Load Balancer ( The NLB is the recommended Load Balancer to be used at Layer 4)
Operates at Layer 4.
Preserves the source IP.
Provides reduced latency compared to other Load Balancers.
Handles millions of requests per seconds.
Third Classic Load Balancer
Operates at Layer 4.
Features Cross Zone Load Balancing.
Idle Connection Timeout:
Is the period of inactivity between the client and the EC2 instance in which the Load Balancer terminates the connection. By default, it is 60 seconds.
Enable Cross Zone Load Balancing, to distribute traffic evenly across the pool of registered AWS instances.
Connection Draining stops the load balancer sending traffic to faulty instances.
Proxy Protocol: To receive the clients IP and User Agent, it needs to be enabled. (only on Network Load Balancer and Classic Load Balancer)
Sticky Sessions: Ensures that the load balancer sends future user request to the EC2 instance that received the initial request.
Auto Scaling Group: Lets you increase or decrease the number of EC2 instances based upon CPU, RAM, Scheduled Scaling (usually used when you predict that there will be an increase in load in X time ex: Marketing Campaign)
Thanks for being here
Comments
Post a Comment