Skip to content

How to Integrate Amazon API Gateway with Amazon EKS? (Terraform & AWS Lambda & Serverless & NLB)

  • You can find the source code for this video in my GitHub Repo.

Intro

In this video, we're going to go over the following sections:

  • Create EKS Cluster Using Terraform
  • Deploy App to Kubernetes and Expose It with NLB
  • Create AWS API Gateway Using Terraform
  • Integrate API Gateway with Amazon EKS

You can find the timestamps in the video description.

Create EKS Cluster Using Terraform

In the first section, let's create an EKS cluster and VPC from scratch.

In the provider, we can declare some version constraints as well as some variables such as the EKS cluster name and an EKS version.

terraform/0-provider.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
terraform {
  required_providers {
    aws = {
      source  = "hashicorp/aws"
      version = "~> 4.0"
    }
  }
}

provider "aws" {
  region = "us-east-1"
}

variable "cluster_name" {
  default = "demo"
}

variable "cluster_version" {
  default = "1.22"
}

Then we need a VPC.

terraform/1-vpc.tf
1
2
3
4
5
6
7
resource "aws_vpc" "main" {
  cidr_block = "10.0.0.0/16"

  tags = {
    Name = "main"
  }
}

Internet gateway.

terraform/2-igw.tf
1
2
3
4
5
6
7
resource "aws_internet_gateway" "igw" {
  vpc_id = aws_vpc.main.id

  tags = {
    Name = "igw"
  }
}

Four subnets, two private and two public.

terraform/3-subnets.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
resource "aws_subnet" "private-us-east-1a" {
  vpc_id            = aws_vpc.main.id
  cidr_block        = "10.0.0.0/19"
  availability_zone = "us-east-1a"

  tags = {
    "Name"                                      = "private-us-east-1a"
    "kubernetes.io/role/internal-elb"           = "1"
    "kubernetes.io/cluster/${var.cluster_name}" = "owned"
  }
}

resource "aws_subnet" "private-us-east-1b" {
  vpc_id            = aws_vpc.main.id
  cidr_block        = "10.0.32.0/19"
  availability_zone = "us-east-1b"

  tags = {
    "Name"                                      = "private-us-east-1b"
    "kubernetes.io/role/internal-elb"           = "1"
    "kubernetes.io/cluster/${var.cluster_name}" = "owned"
  }
}

resource "aws_subnet" "public-us-east-1a" {
  vpc_id                  = aws_vpc.main.id
  cidr_block              = "10.0.64.0/19"
  availability_zone       = "us-east-1a"
  map_public_ip_on_launch = true

  tags = {
    "Name"                                      = "public-us-east-1a"
    "kubernetes.io/role/elb"                    = "1"
    "kubernetes.io/cluster/${var.cluster_name}" = "owned"
  }
}

resource "aws_subnet" "public-us-east-1b" {
  vpc_id                  = aws_vpc.main.id
  cidr_block              = "10.0.96.0/19"
  availability_zone       = "us-east-1b"
  map_public_ip_on_launch = true

  tags = {
    "Name"                                      = "public-us-east-1b"
    "kubernetes.io/role/elb"                    = "1"
    "kubernetes.io/cluster/${var.cluster_name}" = "owned"
  }
}

NAT Gateway to provide internet access for private subnets.

terraform/4-nat.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
resource "aws_eip" "nat" {
  vpc = true

  tags = {
    Name = "nat"
  }
}

resource "aws_nat_gateway" "nat" {
  allocation_id = aws_eip.nat.id
  subnet_id     = aws_subnet.public-us-east-1a.id

  tags = {
    Name = "nat"
  }

  depends_on = [aws_internet_gateway.igw]
}

Routes.

terraform/5-routes.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
resource "aws_route_table" "private" {
  vpc_id = aws_vpc.main.id

  route {
    cidr_block     = "0.0.0.0/0"
    nat_gateway_id = aws_nat_gateway.nat.id
  }

  tags = {
    Name = "private"
  }
}

resource "aws_route_table" "public" {
  vpc_id = aws_vpc.main.id

  route {
    cidr_block = "0.0.0.0/0"
    gateway_id = aws_internet_gateway.igw.id
  }

  tags = {
    Name = "public"
  }
}

resource "aws_route_table_association" "private-us-east-1a" {
  subnet_id      = aws_subnet.private-us-east-1a.id
  route_table_id = aws_route_table.private.id
}

resource "aws_route_table_association" "private-us-east-1b" {
  subnet_id      = aws_subnet.private-us-east-1b.id
  route_table_id = aws_route_table.private.id
}

resource "aws_route_table_association" "public-us-east-1a" {
  subnet_id      = aws_subnet.public-us-east-1a.id
  route_table_id = aws_route_table.public.id
}

resource "aws_route_table_association" "public-us-east-1b" {
  subnet_id      = aws_subnet.public-us-east-1b.id
  route_table_id = aws_route_table.public.id
}

Then the EKS cluster itself.

terraform/6-eks.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
resource "aws_iam_role" "eks-cluster" {
  name = "eks-cluster-${var.cluster_name}"

  assume_role_policy = <<POLICY
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "Service": "eks.amazonaws.com"
      },
      "Action": "sts:AssumeRole"
    }
  ]
}
POLICY
}

resource "aws_iam_role_policy_attachment" "amazon-eks-cluster-policy" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKSClusterPolicy"
  role       = aws_iam_role.eks-cluster.name
}

resource "aws_eks_cluster" "cluster" {
  name     = var.cluster_name
  version  = var.cluster_version
  role_arn = aws_iam_role.eks-cluster.arn

  vpc_config {
    subnet_ids = [
      aws_subnet.private-us-east-1a.id,
      aws_subnet.private-us-east-1b.id,
      aws_subnet.public-us-east-1a.id,
      aws_subnet.public-us-east-1b.id
    ]
  }

  depends_on = [aws_iam_role_policy_attachment.amazon-eks-cluster-policy]
}

Single node group with one node.

terraform/7-nodes.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
resource "aws_iam_role" "nodes" {
  name = "eks-node-group-nodes"

  assume_role_policy = jsonencode({
    Statement = [{
      Action = "sts:AssumeRole"
      Effect = "Allow"
      Principal = {
        Service = "ec2.amazonaws.com"
      }
    }]
    Version = "2012-10-17"
  })
}

resource "aws_iam_role_policy_attachment" "amazon-eks-worker-node-policy" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKSWorkerNodePolicy"
  role       = aws_iam_role.nodes.name
}

resource "aws_iam_role_policy_attachment" "amazon-eks-cni-policy" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKS_CNI_Policy"
  role       = aws_iam_role.nodes.name
}

resource "aws_iam_role_policy_attachment" "amazon-ec2-container-registry-read-only" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly"
  role       = aws_iam_role.nodes.name
}

resource "aws_eks_node_group" "private-nodes" {
  cluster_name    = aws_eks_cluster.cluster.name
  version         = var.cluster_version
  node_group_name = "private-nodes"
  node_role_arn   = aws_iam_role.nodes.arn

  subnet_ids = [
    aws_subnet.private-us-east-1a.id,
    aws_subnet.private-us-east-1b.id
  ]

  capacity_type  = "ON_DEMAND"
  instance_types = ["t3.small"]

  scaling_config {
    desired_size = 1
    max_size     = 5
    min_size     = 0
  }

  update_config {
    max_unavailable = 1
  }

  labels = {
    role = "general"
  }

  depends_on = [
    aws_iam_role_policy_attachment.amazon-eks-worker-node-policy,
    aws_iam_role_policy_attachment.amazon-eks-cni-policy,
    aws_iam_role_policy_attachment.amazon-ec2-container-registry-read-only,
  ]

  # Allow external changes without Terraform plan difference
  lifecycle {
    ignore_changes = [scaling_config[0].desired_size]
  }
}

Now let's create that VPC and a cluster using terraform apply command. 

terraform init
terraform apply

As always, when you use terraform to create a cluster, you need to update your Kubernetes context manually.

aws eks update-kubeconfig --name demo --region us-east-1

Then a quick check to get default service.

kubectl get svc

Deploy App to Kubernetes and Expose It with NLB

Next, we need to deploy an app to Kubernetes and expose it with Network Load Balancer. You don't need to deploy the AWS Load balancer controller for that; you can just use annotations.

k8s/echo-server.yaml
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
---
apiVersion: v1
kind: Namespace
metadata:
  name: staging
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: echoserver
  namespace: staging
spec:
  selector:
    matchLabels:
      app: echoserver
  replicas: 1
  template:
    metadata:
      labels:
        app: echoserver
    spec:
      containers:
      - image: k8s.gcr.io/e2e-test-images/echoserver:2.5
        name: echoserver
        ports:
        - containerPort: 8080
---
apiVersion: v1
kind: Service
metadata:
  name: echoserver
  namespace: staging
  annotations:
    service.beta.kubernetes.io/aws-load-balancer-type: nlb
    service.beta.kubernetes.io/aws-load-balancer-internal: "true"
spec:
  ports:
  - port: 8080
    protocol: TCP
  type: LoadBalancer
  selector:
    app: echoserver

Let's go ahead and apply this app.

kubectl apply -f k8s/echo-server.yaml

If you get services in the staging namespace, you should find the hostname of the load balancer, take a note, or copy it.

kubectl get svc -n staging

Integrate API Gateway with Amazon EKS

The following step is to create AWS API Gateway Using Terraform. We're going to be using API version 2.

terraform/8-api-gateway.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
resource "aws_apigatewayv2_api" "main" {
  name          = "main"
  protocol_type = "HTTP"
}

resource "aws_apigatewayv2_stage" "dev" {
  api_id = aws_apigatewayv2_api.main.id

  name        = "dev"
  auto_deploy = true
}

Go back to the terminal and apply it.

terraform apply

Integrate API Gateway with Amazon EKS

The final step is to integrate API Gateway with Amazon EKS. For integration, we need to create AWS private link, but before, we need to create a security group.

terraform/9-integration.tf
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
resource "aws_security_group" "vpc_link" {
  name   = "vpc-link"
  vpc_id = aws_vpc.main.id

  egress {
    from_port        = 0
    to_port          = 0
    protocol         = "-1"
    cidr_blocks      = ["0.0.0.0/0"]
  }
}

resource "aws_apigatewayv2_vpc_link" "eks" {
  name               = "eks"
  security_group_ids = [aws_security_group.vpc_link.id]
  subnet_ids = [
    aws_subnet.private-us-east-1a.id,
    aws_subnet.private-us-east-1b.id
  ]
}

resource "aws_apigatewayv2_integration" "eks" {
  api_id = aws_apigatewayv2_api.main.id

  integration_uri    = "arn:aws:elasticloadbalancing:us-east-1:<acc-id>:listener/net/a852b4f6ff0be41dfa1505018b083488/e8cf16c1a71e2a37/59bf9fd068f3f993"
  integration_type   = "HTTP_PROXY"
  integration_method = "ANY"
  connection_type    = "VPC_LINK"
  connection_id      = aws_apigatewayv2_vpc_link.eks.id
}

resource "aws_apigatewayv2_route" "get_echo" {
  api_id = aws_apigatewayv2_api.main.id

  route_key = "GET /echo"
  target    = "integrations/${aws_apigatewayv2_integration.eks.id}"
}

output "hello_base_url" {
  value = "${aws_apigatewayv2_stage.dev.invoke_url}/echo"
}

This is the last time when we run terraform.

terraform apply

Use curl to test integration.

curl https://<your-gw-id>.execute-api.us-east-1.amazonaws.com/dev/echo
Back to top