Amazon Linux 2 End of Life 2026: What AWS Users Need to Do

# Check which AMI your instances use
aws ec2 describe-instances \
  --query 'Reservations[].Instances[].{ID:InstanceId,AMI:ImageId,Name:Tags[?Key==`Name`].Value|[0]}' \
  --output table
 
# Check if AMI is Amazon Linux 2
aws ec2 describe-images --image-ids ami-xxx \
  --query 'Images[0].Name'
# "amzn2-ami-hvm-2.0.20240306-x86_64-gp2" ← This is AL2

# ❌ AL2 — EOL June 2026
FROM amazonlinux:2
 
# ✅ AL2023
FROM amazonlinux:2023

data "aws_ami" "al2023" {
  most_recent = true
  owners      = ["amazon"]
 
  filter {
    name   = "name"
    values = ["al2023-ami-*-x86_64"]
  }
 
  filter {
    name   = "virtualization-type"
    values = ["hvm"]
  }
}
 
# For ARM/Graviton instances
data "aws_ami" "al2023_arm" {
  most_recent = true
  owners      = ["amazon"]
 
  filter {
    name   = "name"
    values = ["al2023-ami-*-arm64"]
  }
}

resource "aws_instance" "web" {
  # Before
  # ami = "ami-0abcdef1234567890"  # AL2 AMI
 
  # After
  ami           = data.aws_ami.al2023.id  # AL2023
  instance_type = "t3.micro"
 
  # User data may need updates for AL2023 differences
  user_data = base64encode(templatefile("${path.module}/scripts/setup-al2023.sh", {
    app_version = var.app_version
  }))
}

#!/bin/bash
# AL2 → AL2023 package manager differences
 
# AL2 used amazon-linux-extras
# ❌ amazon-linux-extras install nginx1
# ✅ AL2023 uses dnf directly
dnf install -y nginx
 
# AL2 used yum
# ❌ yum install -y httpd
# ✅ AL2023 uses dnf (yum still works as alias)
dnf install -y httpd
 
# Python
# ❌ AL2: python3 was 3.7
# ✅ AL2023: python3 is 3.9+
python3 --version
 
# SystemD is the same
systemctl enable nginx
systemctl start nginx

resource "aws_lambda_function" "api" {
  function_name = "api-handler"
 
  # Before
  # runtime = "python3.9"    # AL2-based
 
  # After
  runtime = "python3.12"     # AL2023-based
 
  handler  = "handler.main"
  filename = "lambda.zip"
  role     = aws_iam_role.lambda.arn
 
  # Test thoroughly — AL2023 has newer system libraries
}

# Find all Lambda functions using old runtimes
aws lambda list-functions \
  --query 'Functions[?Runtime==`python3.8` || Runtime==`python3.9` || Runtime==`nodejs16.x`].{Name:FunctionName,Runtime:Runtime}' \
  --output table

# Terraform — update all functions
variable "lambda_functions" {
  default = {
    api      = { handler = "api.main" }
    worker   = { handler = "worker.main" }
    notifier = { handler = "notifier.main" }
  }
}
 
resource "aws_lambda_function" "functions" {
  for_each = var.lambda_functions
 
  function_name = each.key
  runtime       = "python3.12"    # AL2023 for all
  handler       = each.value.handler
  # ...
}

resource "aws_eks_node_group" "workers" {
  cluster_name    = aws_eks_cluster.main.name
  node_group_name = "workers"
 
  # Before: AL2 AMI type
  # ami_type = "AL2_x86_64"
 
  # After: AL2023 AMI type
  ami_type       = "AL2023_x86_64_STANDARD"
  instance_types = ["m5.large"]
 
  scaling_config {
    desired_size = 3
    max_size     = 6
    min_size     = 1
  }
 
  # Rolling update — replace nodes gradually
  update_config {
    max_unavailable = 1
  }
}