Browse the Repo

file-type-icon.circleci
file-type-icon_docs
file-type-iconexamples
file-type-iconmodules
file-type-icontest
file-type-icon.gitignore
file-type-iconCODEOWNERS
file-type-iconLICENSE.txt
file-type-iconREADME.md

Browse the Repo

file-type-icon.circleci
file-type-icon_docs
file-type-iconexamples
file-type-iconmodules
file-type-icontest
file-type-icon.gitignore
file-type-iconCODEOWNERS
file-type-iconLICENSE.txt
file-type-iconREADME.md
Elastic Load Balancer (ELB)

Elastic Load Balancer (ELB)

Deploy the Application Load Balancer (ALB) for load balancing HTTP and HTTPS, with support for routing rules and WebSockets.

Preview the Code

mobile file icon

README.md

down

Terraform Version

Load Balancer Modules

This repo contains Gruntwork Modules for running a load balancer in AWS. The modules are:

What about the Classic Load Balancer (CLB) and Network Load Balancer (NLB)?

Note that the Classic Load Balancer (sometimes known by its original name of "ELB") and Network Load Balancers are usually simple enough that it is best defined as part of the Terraform template that needs it, so there are currently no plans to define a standalone "elb" module.

To deploy a classic load balancer, you can use the aws_elb terraform resource.

To deploy a network load balancer, you can use the aws_lb terraform resource with the load_balancer_type input set to "network". You can refer to the after_migration terraform module in our NLB module migration guide for an example of how to configure an NLB using the terraform resources.

Removed modules

The nlb module was removed in version v0.15.0. When Terraform introduced for_each and dynamic sub blocks in v0.12.0, it no longer made sense to maintain the NLB module which thinly wrapped the aws_lb resource to provide dynamic subnet mappings blocks.

Refer to the migration guide for information on how to update your usage.

What is a module?

At Gruntwork, we've taken the thousands of hours we spent building infrastructure on AWS and condensed all that experience and code into pre-built packages or modules. Each module is a battle-tested, best-practices definition of a piece of infrastructure, such as a VPC, ECS cluster, or an Auto Scaling Group. Modules are versioned using Semantic Versioning to allow Gruntwork clients to keep up to date with the latest infrastructure best practices in a systematic way.

How do you use a module?

Most of our modules contain either:

  1. Terraform code
  2. Scripts & binaries

Using a Terraform Module

To use a module in your Terraform templates, create a module resource and set its source field to the Git URL of this repo. You should also set the ref parameter so you're fixed to a specific version of this repo, as the master branch may have backwards incompatible changes (see module sources).

For example, to use v1.0.8 of the ecs-cluster module, you would add the following:

module "ecs_cluster" {
  source = "git::git@github.com:gruntwork-io/module-ecs.git//modules/ecs-cluster?ref=v1.0.8"

  # set the parameters for the ECS cluster module
}

Note: the double slash (//) is intentional and required. It's part of Terraform's Git syntax (see module sources).

See the module's documentation and vars.tf file for all the parameters you can set. Run terraform get -update to pull the latest version of this module from this repo before runnin gthe standard terraform plan and terraform apply commands.

Using scripts & binaries

You can install the scripts and binaries in the modules folder of any repo using the Gruntwork Installer. For example, if the scripts you want to install are in the modules/ecs-scripts folder of the https://github.com/gruntwork-io/module-ecs repo, you could install them as follows:

gruntwork-install --module-name "ecs-scripts" --repo "https://github.com/gruntwork-io/module-ecs" --tag "0.0.1"

See the docs for each script & binary for detailed instructions on how to use them.

Background

What exactly is a load balancer?

A load balancer is usually the public-facing part of your infrastructure, receives incoming requests, makes a decision on which backend service to route them to, and forwards the request.

Load balancers can forward any kind of IP traffic, including TCP, UDP, HTTP, and HTTPS requests.

Why is a load balancer sometimes called a reverse proxy?

A "forward proxy" is a server that forwards incoming requests to an external party on your behalf. For example, a "web proxy" is a forward proxy that makes HTTP requests on your behalf and forwards you the results. This might be useful if you want to hide your IP address from websites you request, or if your organization wants to scan all web traffic.

A "reverse proxy" works in the opposite way. An external party makes an inbound request meant for a server, but instead of hitting that server directly, the request hits the "reverse proxy", which then forwards that request to some unknown backend service, gets the result, and returns it to a client.

It's conceptually identical to a load balancer, so "load balancer" and "reverse proxy" are synonyms.

Why are load balancers needed?

The simplest way to expose a service is to run a single server that runs your service. Users would then make requests directly to this server.

This has the benefit of simplicity, but many drawbacks, too:

  • It lacks "High Availability." If this server goes down, your app loses availability. You could just add more servers, but users tend to access one IP address only.
  • Your app server is in the public subnet, making it a much easier target for hackers.
  • There may be cases when you want to route requests to a service other than this app server.
  • You may want "out of the box" TLS/SSL connections and/or static asset caching, both of which reduce CPU load on your app server.

A load balancer helps achieve all these properties:

  • If a single backend service can fail, the load balancer can just route requests to a healthy one. For this reason, a load balancer usually includes a health check feature to determine if a backend service can receive requests and a way to register multiple backend service instances.
  • A load balancer might sit in a public subnet, but backend services can reside in a private subnet, giving them additional network isolation.
  • A load balancer can inspect the network request it has received and make routing decisions based on its contents.
  • A load balancer can terminate your TLS/SSL connection, cache static assets, and generally reduce load on your app server.

And there are more benefits beyond this. A load balancer can be locked down more than an app server because it doesn't need access to your data stores, it doesn't change very often (i.e. you don't deploy a new version of it every day like you do with app code), and it doesn't typically need file-system access.

Does the load balancer itself need to be High Availability?

Yes! Many clients make network requests using a DNS address like service.acme.com. If that DNS name resolves to a single IP address, then this IP address now becomes a single point of failure.

There are typically two techniques for mitigating this. First, you can make sure that the DNS name service.acme.com resolves to two different IP addresses. Not all network clients will respond to two IP addresses in the same way, but most browsers will immediately try the "second" IP address if the first one fails.

Second, most networking today is Software-Defined Networking, which means that a single IP Address can be dynamically re-assigned to a server. Amazon supports this functionality through the concept of an Elastic IP Address. In fact, an Elastic IP Address can typically be re-assigned to an EC2 Instance in as little as 7 seconds, though Amazon doesn't officially guarantee this fast a transition.

But implementing a dynamic Elastic IP Address failover is highly non-trivial. Fortunately, Amazon's managed load balancers, the Application Load Balancer and Classic Load Balancer, already implement High Availability by including automatic failover under the hood. Amazon's managed load balancers also include auto-scaling for when your load increases, though there are limits to how fast this can happen.

In general, you should not be configuring your own load balancer (e.g. Nginx or HAProxy) unless you can make the case that an existing managed load balancer solution will not meet your needs.

Developing a module

Versioning

We follow the principles of Semantic Versioning. During initial development, the major version is to 0 (e.g., 0.x.y), which indicates the code does not yet have a stable API. Once we hit 1.0.0, we will follow these rules:

  1. Increment the patch version for backwards-compatible bug fixes (e.g., v1.0.8 -> v1.0.9).
  2. Increment the minor version for new features that are backwards-compatible (e.g., v1.0.8 -> 1.1.0).
  3. Increment the major version for any backwards-incompatible changes (e.g. 1.0.8 -> 2.0.0).

The version is defined using Git tags. Use GitHub to create a release, which will have the effect of adding a git tag.

Tests

See the test folder for details.

License

Please see LICENSE.txt for details on how the code in this repo is licensed.

Questions? Ask away.

We're here to talk about our services, answer any questions, give advice, or just to chat.

Ready to hand off the Gruntwork?