DEV Community

Oloruntobi Olurombi
Oloruntobi Olurombi

Posted on

fff

#aws #kubernetes #githubactions #terraform

In the world of cloud-native applications, automation is crucial to streamline deployments and manage infrastructure efficiently. In this article, we’ll explore how to automate the setup of an EKS cluster and deploy a Flask application using Terraform and GitHub Actions. We’ll also touch on security best practices, monitoring, and how to ensure a robust CI/CD pipeline.

Table of Contents

  • Overview

  • Prerequisites

  • Infrastructure Automation: EKS Cluster Setup

  • Application Deployment: Flask App on EKS

  • Monitoring with Prometheus and Grafana

  • Security Best Practices

  • Conclusion

Overview

The goal of this project is to automate the deployment of a containerised Flask application on an EKS (Elastic Kubernetes Service) cluster. Using Terraform to provision AWS resources and GitHub Actions to automate the CI/CD pipeline, this setup allows for seamless infrastructure management and application deployment.

Why Terraform?

Terraform enables you to write declarative code for infrastructure. Instead of manually creating resources like VPCs, subnets, or an EKS cluster, we automate everything via Infrastructure as Code (IaC).

Why GitHub Actions?

GitHub Actions provides a powerful way to integrate CI/CD, testing, static analysis, and security checks into the code deployment process.

Prerequisites

Before diving into the automation, here are the prerequisites you’ll need to get started:

  • AWS Account: Create an AWS account if you don’t have one.

  • IAM Access Keys: Set up access keys with permissions for managing EKS, EC2, and S3.

  • S3 Bucket: Create an S3 bucket to store your Terraform state files securely.

  • AWS CLI: Install and configure the AWS CLI.

  • Terraform: Make sure Terraform is installed on your local machine or use GitHub Actions for automation.

  • GitHub Secrets: Add AWS credentials (access keys, secret keys) and other sensitive data as GitHub secrets to avoid hardcoding them.

  • Synk: Create a Synk account and get your Token.

  • SonarCloud: Create a SonarCloud account and get your Token, Organisation key and Project key.

Infrastructure Automation: EKS Cluster Setup

Automating infrastructure deployment is key to maintaining scalable, consistent, and reliable environments. In this project, Terraform is utilised to automate the provisioning of an EKS cluster, its node groups, and the supporting AWS infrastructure. This includes VPC creation, IAM roles, S3 bucket setup, and cloud resources like CloudWatch and CloudTrail for logging and monitoring.

Terraform Setup

Let’s start by provisioning the necessary infrastructure. Below is the detailed explanation of the key resources defined in the Terraform files.

EKS Cluster and Node Group (main.tf):

This provision an EKS cluster and node group with IAM roles attached.
The cluster supports encryption using a KMS key, and the worker nodes are set up to scale between a minimum of 2 nodes. Outputs include the cluster name and endpoint for easy reference.

touch main.tf
Enter fullscreen mode Exit fullscreen mode 
terraform {
  backend "s3" {
    bucket = "regtech-iac"
    key = "terraform.tfstate"
    region = "us-east-1"
    encrypt = true 
  }
}

# Provides an EKS Cluster
resource "aws_eks_cluster" "eks_cluster" {
  name     = var.cluster_name
  role_arn = aws_iam_role.eks_cluster_role.arn

  version = "1.28"

  vpc_config {
    subnet_ids = [aws_subnet.public_subnet_1.id, aws_subnet.public_subnet_2.id, aws_subnet.public_subnet_3.id]
  }

  encryption_config {
    provider {
      key_arn = aws_kms_key.eks_encryption_key.arn
    }
    resources = ["secrets"]
  }

  # Ensure that IAM Role permissions are created before and deleted after EKS Cluster handling.
  # Otherwise, EKS will not be able to properly delete EKS managed EC2 infrastructure such as Security Groups.
  depends_on = [
    aws_iam_role_policy_attachment.eks_cluster_policy_attachment,
    aws_iam_role_policy_attachment.eks_service_policy_attachment,
  ]
}

# Provides an EKS Node Group 

resource "aws_eks_node_group" "eks_node_group" {
  cluster_name    = aws_eks_cluster.eks_cluster.name
  node_group_name = var.node_group_name
  node_role_arn   = aws_iam_role.eks_node_group_role.arn
  subnet_ids      = [aws_subnet.public_subnet_1.id, aws_subnet.public_subnet_2.id, aws_subnet.public_subnet_3.id]

  scaling_config {
    desired_size = 2
    max_size     = 2
    min_size     = 2
  }
  update_config {
    max_unavailable = 1
  }

  # Ensure that IAM Role permissions are created before and deleted after EKS Node Group handling.
  # Otherwise, EKS will not be able to properly delete EC2 Instances and Elastic Network Interfaces.
  depends_on = [
    aws_iam_role_policy_attachment.eks_worker_node_policy_attachment,
    aws_iam_role_policy_attachment.eks_cni_policy_attachment,
    aws_iam_role_policy_attachment.ec2_container_registry_readonly,
  ]
}

# Extra resources 
resource "aws_ebs_volume" "volume_regtech"{
    availability_zone = var.az_a
    size = 40
    encrypted = true
    type = "gp2"
    kms_key_id        = aws_kms_key.ebs_encryption_key.arn
}

resource "aws_s3_bucket" "regtech_iac" {
  bucket = var.bucket_name
}

resource "aws_s3_bucket_server_side_encryption_configuration" "regtech_iac_encrypt_config" {
    bucket = aws_s3_bucket.regtech_iac.bucket
    rule {
        apply_server_side_encryption_by_default {
        kms_master_key_id = aws_kms_key.s3_encryption_key.arn  
        sse_algorithm = "aws:kms"
        }
    }
}


# OutPut Resources
output "endpoint" {
  value = aws_eks_cluster.eks_cluster.endpoint
}

output "eks_cluster_name" {
    value = aws_eks_cluster.eks_cluster.name
}
Enter fullscreen mode Exit fullscreen mode
Networking (vpc.tf):

Defines a VPC, public subnets for the EKS cluster, and private subnets for other resources, ensuring flexibility in network architecture.

vpc.tf
Enter fullscreen mode Exit fullscreen mode 
# Provides a VPC resource
resource "aws_vpc" "main" {
  cidr_block       = var.vpc_cidr_block
  instance_tenancy = "default"

  tags = {
    Name = var.tags_vpc
  }
}


# Provides an VPC Public subnet resource
resource "aws_subnet" "public_subnet_1" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.p_s_1_cidr_block
  availability_zone = var.az_a
  map_public_ip_on_launch = true

  tags = {
    Name = var.tags_public_subnet_1
  }
}

resource "aws_subnet" "public_subnet_2" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.p_s_2_cidr_block
  availability_zone = var.az_b
  map_public_ip_on_launch = true

  tags = {
    Name = var.tags_public_subnet_2
  }
}

resource "aws_subnet" "public_subnet_3" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.p_s_3_cidr_block
  availability_zone = var.az_c
  map_public_ip_on_launch = true

  tags = {
    Name = var.tags_public_subnet_3
  }
}

# Provides an VPC Private subnet resource
resource "aws_subnet" "private_subnet_1" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.private_s_1_cidr_block
  availability_zone = var.az_private_a
  map_public_ip_on_launch = false 

  tags = {
    Name = var.tags_private_subnet_1
  }
}

resource "aws_subnet" "private_subnet_2" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.private_s_2_cidr_block
  availability_zone = var.az_private_b
  map_public_ip_on_launch = false 

  tags = {
    Name = var.tags_private_subnet_2
  }
}

resource "aws_subnet" "private_subnet_3" {
  vpc_id     = aws_vpc.main.id
  cidr_block = var.private_s_3_cidr_block
  availability_zone = var.az_private_c
  map_public_ip_on_launch = false 

  tags = {
    Name = var.tags_private_subnet_3
  }
}
Enter fullscreen mode Exit fullscreen mode
IAM Roles (iam.tf):

IAM roles and policies for the EKS cluster, node groups, and autoscaler. Includes roles for security services like CloudWatch and CloudTrail, ensuring robust monitoring.

iam.tf
Enter fullscreen mode Exit fullscreen mode 
# Declare the aws_caller_identity data source
data "aws_caller_identity" "current" {}


# IAM Role for EKS Cluster Plane 

resource "aws_iam_role" "eks_cluster_role" {
    name = var.eks_cluster_role_name

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

resource "aws_iam_role_policy_attachment" "eks_cluster_policy_attachment" {
    policy_arn = "arn:aws:iam::aws:policy/AmazonEKSClusterPolicy"
    role = aws_iam_role.eks_cluster_role.name 
}

resource "aws_iam_role_policy_attachment" "eks_service_policy_attachment" {
  role       = aws_iam_role.eks_cluster_role.name
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKSServicePolicy"
}

# IAM Role for Worker node

resource "aws_iam_role" "eks_node_group_role" {
    name = var.eks_node_group_role_name

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

resource "aws_iam_role_policy_attachment" "eks_worker_node_policy_attachment" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKSWorkerNodePolicy"
  role       = aws_iam_role.eks_node_group_role.name
}

resource "aws_iam_role_policy_attachment" "eks_cni_policy_attachment" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEKS_CNI_Policy"
  role       = aws_iam_role.eks_node_group_role.name
}

resource "aws_iam_role_policy_attachment" "ec2_container_registry_readonly" {
  policy_arn = "arn:aws:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly"
  role       = aws_iam_role.eks_node_group_role.name
}

resource "aws_iam_instance_profile" "eks_node_instance_profile" {
    name = var.eks_node_group_profile
    role = aws_iam_role.eks_node_group_role.name
}


# Policy For volume creation and attachment

resource "aws_iam_role_policy" "eks_node_group_volume_policy" {
  name   = var.eks_node_group_volume_policy_name
  role   = aws_iam_role.eks_node_group_role.name
  policy = jsonencode({
    "Version": "2012-10-17",
    "Statement": [
      {
        "Effect": "Allow",
        "Action": [
          "ec2:CreateTags",
          "ec2:DescribeTags",
          "ec2:DescribeVolumes",
          "ec2:DescribeVolumeStatus",
          "ec2:CreateVolume",
          "ec2:AttachVolume"
        ],
        "Resource": "arn:aws:ec2:${var.region}:${data.aws_caller_identity.current.account_id}:volume/*"
      }
    ]
  })
}


# IAM Role for CloudWatch 

resource "aws_iam_role" "cloudwatch_role" {
  name = "cloudwatch_role_log"

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

resource "aws_iam_role_policy_attachment" "cloudwatch_policy_attachment" {
  role       = aws_iam_role.cloudwatch_role.name
  policy_arn = "arn:aws:iam::aws:policy/CloudWatchLogsFullAccess"
}

# IAM Role for CloudTrail

resource "aws_iam_role" "cloudtrail_role" {
  name = "cloudtrail_role_log"

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

resource "aws_iam_role_policy_attachment" "cloudtrail_policy_attachment" {
  role       = aws_iam_role.cloudtrail_role.name
  policy_arn = "arn:aws:iam::aws:policy/AWSCloudTrail_FullAccess"
}


# KMS Key Policy for Encryption

resource "aws_kms_key" "ebs_encryption_key" {
  description = "KMS key for EBS volume encryption"
}

resource "aws_kms_key" "s3_encryption_key" {
  description = "KMS key for S3 bucket encryption"
}

resource "aws_kms_key" "eks_encryption_key" {
  description = "KMS key for EKS secret encryption"
}


resource "aws_s3_bucket_policy" "regtech_iac_policy" {
  bucket = aws_s3_bucket.regtech_iac.id

  policy = jsonencode({
    Version = "2012-10-17"
    Statement = [
      {
        Effect = "Allow"
        Principal = {
          Service = "cloudtrail.amazonaws.com"
        }
        Action = "s3:GetBucketAcl"
        Resource = "arn:aws:s3:::${aws_s3_bucket.regtech_iac.bucket}"
      },
      {
        Effect = "Allow"
        Principal = {
          Service = "cloudtrail.amazonaws.com"
        }
        Action = "s3:PutObject"
        Resource = "arn:aws:s3:::${aws_s3_bucket.regtech_iac.bucket}/AWSLogs/${data.aws_caller_identity.current.account_id}/*"
        Condition = {
          StringEquals = {
            "s3:x-amz-acl" = "bucket-owner-full-control"
          }
        }
      }
    ]
  })
}
Enter fullscreen mode Exit fullscreen mode
CloudWatch and Monitoring (cloudwatch.tf):

This provisions CloudWatch log groups, an SNS topic for alerts, and a CloudWatch alarm to monitor CPU utilisation. CloudTrail logs are configured to monitor S3 and management events.

touch cloudwatch.tf
Enter fullscreen mode Exit fullscreen mode 
resource "aws_cloudwatch_log_group" "eks_log_group" {
  name              = "/aws/eks/cluster-logs-regtech"
  retention_in_days = 30
}

resource "aws_cloudtrail" "security_trail" {
  name                          = "security-trail-log"
  s3_bucket_name                = aws_s3_bucket.regtech_iac.bucket
  include_global_service_events = true
  is_multi_region_trail         = true
  enable_log_file_validation    = true

  event_selector {
    read_write_type           = "All"
    include_management_events = true

    data_resource {
      type = "AWS::S3::Object"
      values = ["arn:aws:s3:::${aws_s3_bucket.regtech_iac.bucket}/"]
    }
  }
}


resource "aws_sns_topic" "alarm_topic" {
    name = "high-cpu-alarm-topic"
}

resource "aws_sns_topic_subscription" "alarm_subscription" {
    topic_arn = aws_sns_topic.alarm_topic.arn
    protocol = "email"
    endpoint = "oloruntobiolurombi@gmail.com"
}

resource "aws_cloudwatch_metric_alarm" "cpu_alarm" {
  alarm_name          = "high_cpu_usage"
  comparison_operator = "GreaterThanThreshold"
  evaluation_periods  = "2"
  metric_name         = "CPUUtilization"
  namespace           = "AWS/EC2"
  period              = "120"
  statistic           = "Average"
  threshold           = "70"

  alarm_actions = [
    aws_sns_topic.alarm_topic.arn
  ]
}
Enter fullscreen mode Exit fullscreen mode
AutoScaler IAM (iam-autoscaler.tf):

This will provision roles and policies for enabling the EKS Cluster Autoscaler are included, which will help in adjusting the number of worker nodes based on resource demands.

data "aws_iam_policy_document" "eks_cluster_autoscaler_assume_role_policy" {
  statement {
    actions = ["sts:AssumeRoleWithWebIdentity"]
    effect  = "Allow"

    condition {
      test     = "StringEquals"
      variable = "${replace(aws_iam_openid_connect_provider.eks.url, "https://", "")}:sub"
      values   = ["system:serviceaccount:kube-system:cluster-autoscaler"]
    }

    principals {
      identifiers = [aws_iam_openid_connect_provider.eks.arn]
      type        = "Federated"
    }
  }
}

resource "aws_iam_role" "eks_cluster_autoscaler" {
  assume_role_policy = data.aws_iam_policy_document.eks_cluster_autoscaler_assume_role_policy.json
  name               = "eks-cluster-autoscaler"
}

resource "aws_iam_policy" "eks_cluster_autoscaler" {
  name = "eks-cluster-autoscaler"

  policy = jsonencode({
    Statement = [{
      Action = [
                "autoscaling:DescribeAutoScalingGroups",
                "autoscaling:DescribeAutoScalingInstances",
                "autoscaling:DescribeLaunchConfigurations",
                "autoscaling:DescribeTags",
                "autoscaling:SetDesiredCapacity",
                "autoscaling:TerminateInstanceInAutoScalingGroup",
                "ec2:DescribeLaunchTemplateVersions"
            ]
      Effect   = "Allow"
      Resource = "*"
    }]
    Version = "2012-10-17"
  })
}

resource "aws_iam_role_policy_attachment" "eks_cluster_autoscaler_attach" {
  role       = aws_iam_role.eks_cluster_autoscaler.name
  policy_arn = aws_iam_policy.eks_cluster_autoscaler.arn
}

output "eks_cluster_autoscaler_arn" {
  value = aws_iam_role.eks_cluster_autoscaler.arn
}
Enter fullscreen mode Exit fullscreen mode
Routing (security_groups.tf)

This defines the security groups required for your infrastructure. Security groups act as virtual firewalls that control the inbound and outbound traffic to your resources.

touch security_groups.tf
Enter fullscreen mode Exit fullscreen mode 
# Provides a security group 
resource "aws_security_group" "main_sg" {
    name = "main_sg"
    description = var.main_sg_description
    vpc_id = aws_vpc.main.id 

    ingress  {
        description = "ssh access"
        from_port = 22
        to_port = 22
        protocol = "tcp"
        cidr_blocks = ["0.0.0.0/0"]
    }

    ingress  {
        description = "Kubernetes API access"
        from_port = 443
        to_port = 443 
        protocol = "tcp"
        cidr_blocks = ["0.0.0.0/0"]
    }

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

    tags = {
        Name = var.tags_main_sg_eks
    }
}
Enter fullscreen mode Exit fullscreen mode
Variables (variables.tf)

This contains the variable definitions used across your Terraform configurations. These variables provide default values and can be overridden as needed.

touch variables.tf
Enter fullscreen mode Exit fullscreen mode 
variable "region" {
    type = string 
    default = "us-east-1"
}

#variable "profile" {
#  type = string
#  default = "tobi"
#}

variable "bucket_name" {
    type = string 
    default = "regtech-logs"
}

variable "aws_access_key_id" {
    type = string
    default = ""
}

variable "aws_secret_access_key" {
    type = string
    default = ""
}

variable "tags_vpc" {
    type = string 
    default = "main-vpc-eks"
}

variable "tags_public_rt" {
    type = string 
    default = "public-route-table"
}

variable "tags_igw" {
    type = string 
    default = "internet-gateway"
}

variable "tags_public_subnet_1" {
    type = string 
    default = "public-subnet-1"
}

variable "tags_public_subnet_2" {
    type = string 
    default = "public-subnet-2"
}

variable "tags_public_subnet_3" {
    type = string 
    default = "public-subnet-3"
}

variable "tags_private_subnet_1" {
    type = string 
    default = "private-subnet-1"
}

variable "tags_private_subnet_2" {
    type = string 
    default = "private-subnet-2"
}

variable "tags_private_subnet_3" {
    type = string 
    default = "private-subnet-3"
}

variable "tags_main_sg_eks" {
    type = string
    default = "main-sg-eks"
}

variable "instance_type" {
    type = string 
    default = "t2.micro"
}

variable "cluster_name" {
    type = string 
    default = "EKSCluster"
}

variable "node_group_name" {
    type = string 
    default = "SlaveNode"
}

variable "vpc_cidr_block" {
    type = string 
    default = "10.0.0.0/16"
}

variable "p_s_1_cidr_block" {
    type = string 
    default = "10.0.1.0/24"
}

variable "az_a" {
    type = string 
    default = "us-east-1a"
}

variable "p_s_2_cidr_block" {
    type = string 
    default = "10.0.2.0/24"
}

variable "az_b" {
    type = string 
    default = "us-east-1b"
}

variable "p_s_3_cidr_block" {
    type = string 
    default = "10.0.3.0/24"
}

variable "az_c" {
    type = string 
    default = "us-east-1c"
}

variable "private_s_1_cidr_block" {
    type = string 
    default = "10.0.4.0/24"
}

variable "az_private_a" {
    type = string 
    default = "us-east-1c"
}

variable "private_s_2_cidr_block" {
    type = string 
    default = "10.0.5.0/24"
}

variable "az_private_b" {
    type = string 
    default = "us-east-1c"
}

variable "private_s_3_cidr_block" {
    type = string 
    default = "10.0.6.0/24"
}

variable "az_private_c" {
    type = string 
    default = "us-east-1c"
}

variable "main_sg_description" {
    type = string 
    default = "Allow TLS inbound traffic and all outbound traffic"
}


variable "eks_node_group_profile" {
    type = string 
    default = "eks-node-group-instance-profile_log"
}

variable "eks_cluster_role_name" {
    type = string 
    default = "eksclusterrole_log"
}

variable "eks_node_group_role_name" {
    type = string 
    default = "eks-node-group-role_log"
}

variable "eks_node_group_volume_policy_name" {
    type = string 
    default = "eks-node-group-volume-policy"
}

variable "eks_describe_cluster_policy_name" {
    type = string 
    default = "eks-describe-cluster-policy_log"
}

variable "tags_nat" {
    type = string 
    default = "nat-gateway_eip"
}

variable "tags_k8s-nat" {
    type = string 
    default = "k8s-nat"
}
Enter fullscreen mode Exit fullscreen mode
Provider (provider.tf)

This is crucial in any Terraform project as it defines the provider configuration, which in this case is AWS.

touch
Enter fullscreen mode Exit fullscreen mode 
terraform {
  required_providers {
    aws = {
      source  = "hashicorp/aws"
      version = "~> 5.0"
    }
  }
}

# Configure the AWS Provider
provider "aws" {
  region = var.region
  access_key = var.aws_access_key_id
  secret_key = var.aws_secret_access_key 
}
Enter fullscreen mode Exit fullscreen mode
IAM OpenID (oidc.tf)

This provides an IAM OpenID Connect provider.

touch oidc.tf
Enter fullscreen mode Exit fullscreen mode 
data "tls_certificate" "eks" {
  url = aws_eks_cluster.eks_cluster.identity[0].oidc[0].issuer
}

resource "aws_iam_openid_connect_provider" "eks" {
  client_id_list  = ["sts.amazonaws.com"]
  thumbprint_list = [data.tls_certificate.eks.certificates[0].sha1_fingerprint]
  url             = aws_eks_cluster.eks_cluster.identity[0].oidc[0].issuer
}
Enter fullscreen mode Exit fullscreen mode

Top comments (0)

Read next

ktauchathuranga profile image

Automating Arduino Library Deployment with GitHub Actions: Version Validation, Pull Requests, and Release Automation

Ashen Chathuranga -

imsushant12 profile image

Conclusion of My Node.js Journey and a Sneak Peek into My Upcoming AWS Series

Sushant Gaurav -

ealtinor profile image

How to Generate Prometheus Metrics from Logs Collected by Fluentd

Enes Altınorak -

mrcaption49 profile image

Docker | Docker Compose | Kubernetes

Pranav Bakare -