How Elastic Load Balancing Works
A load balancer accepts incoming traffic from clients and routes requests to its registered EC2 instances in one or more Availability Zones. The load balancer also monitors the health of its registered instances and ensures that it routes traffic only to healthy instances. When the load balancer detects an unhealthy instance, it stops routing traffic to that instance, and then resumes routing traffic to that instance when it detects that the instance is healthy again.
You configure your load balancer to accept incoming traffic by specifying one or more listeners. A listener is a process that checks for connection requests. It is configured with a protocol and port number for connections from clients to the load balancer and a protocol and port number for connections from the load balancer to the instances.
Elastic Load Balancing supports two types of load balancers: Application Load Balancers and Classic Load Balancers. There is a key difference between the way you configure these load balancers. With a Classic Load Balancer, you register instances with the load balancer. With an Application Load Balancer, you register the instances as targets in a target group, and route traffic to a target group.
Availability Zones and Instances
When you enable an Availability Zone for your load balancer, Elastic Load Balancing creates a load balancer node in the Availability Zone. If you register instances in an Availability Zone but do not enable the Availability Zone, these registered instances do not receive traffic.
With a Classic Load Balancer, we recommend that you enable multiple Availability Zones. With an Application Load Balancer, we require you to enable multiple Availability Zones. After you enable multiple Availability Zones, if one Availability Zone becomes unavailable or has no healthy instances, the load balancer can continue to route traffic to the healthy registered instances in another Availability Zone.
Cross-zone load balancing is always enabled for an Application Load Balancer and is disabled by default for a Classic Load Balancer. If cross-zone load balancing is enabled, the load balancer distributes traffic evenly across all registered instances in all enabled Availability Zones. If cross-zone load balancing is disabled, the load balancer distributes traffic evenly across all enabled Availability Zones. For example, suppose that you have 10 instances in Availability Zone us-west-2a and 2 instances in us-west-2b. If cross-zone load balancing is disabled, the requests are distributed evenly between us-west-2a and us-west-2b. As a result, the 2 instances in us-west-2b serve the same amount of traffic as the 10 instances in us-west-2a. However, if cross-zone load balancing is enabled, the load balancer distributes incoming requests evenly across all 12 instances.
Before a client sends a request to your load balancer, it resolves the load balancer's
domain name using a Domain Name System (DNS) server. The DNS entry is controlled by
Amazon, because your instances are in the
amazonaws.com domain. The Amazon
DNS servers return one or more IP addresses to the client, which are the IP addresses of
the load balancer nodes for your load balancer. As traffic to your application changes
over time, Elastic Load Balancing scales your load balancer and updates the DNS entry. Note that the DNS
entry also specifies the time-to-live (TTL) as 60 seconds, which ensures that the IP
addresses can be remapped quickly in response to changing traffic.
The client determines which IP address to use to send requests to the load balancer. The load balancer node that receives the request selects a healthy registered instance and sends the request to the instance using its private IP address.
With a Classic Load Balancer, the load balancer node that receives the request selects a registered instance using the round robin routing algorithm for TCP listeners and the least outstanding requests routing algorithm for HTTP and HTTPS listeners.
With an Application Load Balancer, the load balancer node that receives the request selects a registered target from the target group using the round robin routing algorithm. Routing is performed independently for each target group, even when a target is registered with multiple target groups.
Classic Load Balancers use pre-open connections but Application Load Balancers do not.
Classic Load Balancers support the following protocols on front-end connections (client to load balancer): HTTP/0.9, HTTP/1.0, and HTTP/1.1.
Application Load Balancers support the following protocols on front-end connections: HTTP/0.9, HTTP/1.0, HTTP/1.1, and HTTP/2. You can use HTTP/2 only with HTTPS listeners, and send up to 128 requests in parallel using one HTTP/2 connection. Application Load Balancers also support connection upgrades from HTTP to Websockets.
Both Classic Load Balancers and Application Load Balancers use HTTP/1.1 on back-end connections (load balancer to registered instance). Keep-alive is supported on back-end connections by default. For HTTP/1.0 requests from clients that do not have a host header, the load balancer generates a host header for the HTTP/1.1 requests sent on the back-end connections. For Classic Load Balancer, the host header contains the IP address of the load balancer node. For Application Load Balancer, the host header contains the DNS name of the load balancer.
You can set an idle timeout value for both Classic Load Balancers and Application Load Balancers. The default value is 60 seconds. With a Classic Load Balancer, if a front-end connection or a back-end connection is idle for longer than the idle timeout value, the connection is torn down and the client receives an error response. With an Application Load Balancer, the idle timeout value applies only to front-end connections. A registered target can use a keep-alive timeout to keep a back-end connection open until it is ready to tear it down.
Classic Load Balancers and Application Load Balancers support pipelined HTTP on front-end connections. They do not support pipelined HTTP on back-end connections.
Classic Load Balancers and Application Load Balancers support X-Forwarded-For, X-Forwarded-Proto, and X-Forwarded-Port headers.
Classic Load Balancers and Application Load Balancers honor the connection header from the incoming client request after proxying the response back to the client.
HTTP headers for Application Load Balancers have the following size limits:
Request line: 8K
Single header: 8K
Whole header: 64K
Load Balancer Scheme
When you create a load balancer, you must choose whether to make it an internal load balancer or an Internet-facing load balancer. Note that when you create a Classic Load Balancer in EC2-Classic, it must be an Internet-facing load balancer.
The nodes of an Internet-facing load balancer have public IP addresses. The DNS name of an Internet-facing load balancer is publicly resolvable to the public IP addresses of the nodes. Therefore, Internet-facing load balancers can route requests from clients over the Internet.
The nodes of an internal load balancer have only private IP addresses. The DNS name of an internal load balancer is publicly resolvable to the private IP addresses of the nodes. Therefore, internal load balancers can only route requests from clients with access to the VPC for the load balancer.
Note that both Internet-facing and internal load balancers route requests to your instances using private IP addresses. Therefore, your instances do not need public IP addresses to receive requests from an internal or an Internet-facing load balancer.
If your application has multiple tiers, for example web servers that must be connected to the Internet and database servers that are only connected to the web servers, you can design an architecture that uses both internal and Internet-facing load balancers. Create an Internet-facing load balancer and register the web servers with it. Create an internal load balancer and register the database servers with it. The web servers receive requests from the Internet-facing load balancer and send requests for the database servers to the internal load balancer. The database servers receive requests from the internal load balancer.