Containers

Optimize IP addresses usage by pods in your Amazon EKS cluster

Many enterprise customers adopt multi-account strategy to meet their business needs and at the same time reduce the security blast radius. Customers have had problems maintaining network topology because of constant growth and increased workloads. They can quickly run out of IP space while planning out the VPC Classless Inter-Domain Routing (CIDR). In this blog, we briefly explain two solution options to address this problem, one using custom networking and another without.

Things you should know

Kubernetes: Kubernetes is a popular open source orchestration system of deploying and managing containerized applications, at scale.

Amazon EKS: Amazon Elastic Kubernetes Service (Amazon EKS) is a managed service that makes it easy for you to run Kubernetes on AWS without needing to stand up or maintain your own infrastructure. This fully managed service is designed to provide scalability, security, and  integrations with other AWS services like Amazon Elastic Container Registry (Amazon ECR), IAM, and Elastic Load Balancer, to name a few.

AWS VPC CNI plugin: The AWS VPC CNI networking plugin provides native networking capabilities for pods running inside a Virtual Private Cloud (VPC). This plugin is responsible for assigning IPs to the pods whenever the pods are created. In order to assign an IP, the plugin needs to be aware of the underlying instance’s ENI capacity and know how many primary and secondary interfaces are supported by the particular instance type.

Customer scenarios:

There are customers with tens or hundreds of accounts and thousands of workloads. As customers scale the EKS clusters and deploy more workloads, the number of pods managed by a cluster can grow easily above thousands of pods. Now, each of these pods will consume an IP address. This scenario might become challenging as the availability of IP addresses on a VPC is limited and it’s not always possible to recreate a larger VPC or extend the current VPC’s CIDR blocks. It’s worth noting that both the worker nodes and the pods themselves require IP addresses. Also, the CNI will reserve IP addresses for future use.

Now, if you have pods executing in your EKS cluster and are already running out of IP addresses on your primary VPC CIDR (e.g. 10.0.0.0/20), we have two options:

  • CNI custom networking: this option uses the custom networking plugin to maintain your workers nodes’ IP addresses on your primary VPC CIDR (in this example, 10.0.0.0/20), but move your pods’ IP addresses to a larger subnet (e.g. 100.64.0.0/8). In this scenario, you can move all your pods to the new range and still use your 10.0.0.0/20 CIDR IP addresses as the source IP. However, here are some considerations you should know while using this configuration:
    • This solution will result in getting lesser IP addresses for the pods.
    • The solution is comparatively complex as it involves manual calculation & configuration of “max pods.”
    • The solution is not supported by EKS managed node groups.
  • Secondary CIDR with VPC: another option is to deploy new worker nodes with both the instance and pods networking on a new larger CIDR block (e.g. 100.64.0.0/8). In this scenario, after adding the new CIDR to your VPC, you can deploy another node group using the secondary CIDR and drain the original nodes to automatically redeploy the pods onto the new worker nodes.

Option 1: CNI custom networking

As shown in the diagram below, the primary Elastic Network Interface (ENI) of the worker node still uses the primary VPC CIDR range (in this case 10.0.0.0/20) but the secondary ENIs for pods use the secondary VPC CIDR Range (in this case 100.64.0.0/10). Now, in order to have the pods using the 100.64.0.0/8 CIDR range, you will have to configure the CNI plugin to use custom networking. You can follow through the steps as documented here. You can also use a script published in this blog to execute and configure the plugin effortlessly.

Figure 1 - Architecture reference for option 1

Option 2: Unique network for nodes and pods.

In this scenario, you can still maintain the EKS cluster as is (the control plane will still live on the original subnet/s) but you’ll completely migrate the instances and the pods to a secondary subnet/s. In the example shown on the diagram below, both the primary and the secondary ENIs use the secondary VPC CIDR Range (in this case 100.64.0.0/10).

Figure 2 - Architecture reference option 2

How to use secondary CIDR with EKS (Option 2):

The following steps help you set up an Amazon EKS cluster and the managed worker nodes (data plane), followed by steps to drain worker nodes and move the pods to new worker nodes that you planned to retain. We have used Amazon AWS CLI to execute all the steps, which can also be easily automated in your CICD pipeline. You can execute these steps from your local machine, by logging into an Amazon EC2 instance or AWS SSM Session Manager.

Step 1: let’s begin by creating Amazon Virtual Private Cloud (VPC) and subnets where we can deploy Amazon EKS clusters. If you plan on using an existing VPC, you can skip this step and directly jump to step 2.

Set the environment variables like CLUSTER_NAME and KEY_NAME that can used for subsequent commands

export CLUSTER_NAME=eks-cluster
export KEY_NAME=<an existing Keypair>

This CloudFormation stack will deploy a VPC and private subnets on the 10.0.0.0/20 CIDR block. Save the below CloudFormation script as eks-vpc.yaml and execute the following AWS CLI command to deploy the stack.

AWSTemplateFormatVersion: '2010-09-09'

Description:  This template deploys a VPC, with three public and private subnets spread
  across three Availability Zones. It deploys an internet gateway, with a default
  route on the public subnets. It deploys three NAT gateways (one in each AZ),
  and default routes for them in the private subnets.

Parameters:
  EnvironmentName:
    Description: An environment name that is prefixed to resource names
    Type: String
  VpcCIDR:
    Description: Please enter the IP range (CIDR notation) for this VPC
    Type: String
    Default: 10.0.0.0/16
  PublicSubnet1CIDR:
    Description: Please enter the IP range (CIDR notation) for the public subnet in the first Availability Zone
    Type: String
    Default: 10.0.0.0/20
  PublicSubnet2CIDR:
    Description: Please enter the IP range (CIDR notation) for the public subnet in the second Availability Zone
    Type: String
    Default: 10.0.16.0/20
  PublicSubnet3CIDR:
    Description: Please enter the IP range (CIDR notation) for the public subnet in the second Availability Zone
    Type: String
    Default: 10.0.32.0/20
  PrivateSubnet1CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the first Availability Zone
    Type: String
    Default: 10.0.48.0/20
  PrivateSubnet2CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the second Availability Zone
    Type: String
    Default: 10.0.64.0/20
  PrivateSubnet3CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the second Availability Zone
    Type: String
    Default: 10.0.80.0/20

Resources:
  VPC:
    Type: AWS::EC2::VPC
    Properties:
      CidrBlock: !Ref VpcCIDR
      EnableDnsSupport: true
      EnableDnsHostnames: true
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-vpc

  InternetGateway:
    Type: AWS::EC2::InternetGateway
    Properties:
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-igw

  InternetGatewayAttachment:
    Type: AWS::EC2::VPCGatewayAttachment
    Properties:
      InternetGatewayId: !Ref InternetGateway
      VpcId: !Ref VPC

  PublicSubnet1:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 0, !GetAZs '' ]
      CidrBlock: !Ref PublicSubnet1CIDR
      MapPublicIpOnLaunch: true
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-public-subnet-1
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PublicSubnet2:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 1, !GetAZs  '' ]
      CidrBlock: !Ref PublicSubnet2CIDR
      MapPublicIpOnLaunch: true
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-public-subnet-2
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PublicSubnet3:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 2, !GetAZs  '' ]
      CidrBlock: !Ref PublicSubnet3CIDR
      MapPublicIpOnLaunch: true
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-public-subnet-3
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateSubnet1:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 0, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet1CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-subnet-1
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateSubnet2:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 1, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet2CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-subnet-2
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateSubnet3:
    Type: AWS::EC2::Subnet
    Properties:
      VpcId: !Ref VPC
      AvailabilityZone: !Select [ 2, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet3CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-subnet-3
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  NatGateway1EIP:
    Type: AWS::EC2::EIP
    DependsOn: InternetGatewayAttachment
    Properties:
      Domain: vpc

  NatGateway2EIP:
    Type: AWS::EC2::EIP
    DependsOn: InternetGatewayAttachment
    Properties:
      Domain: vpc

  NatGateway3EIP:
    Type: AWS::EC2::EIP
    DependsOn: InternetGatewayAttachment
    Properties:
      Domain: vpc

  NatGateway1:
    Type: AWS::EC2::NatGateway
    Properties:
      AllocationId: !GetAtt NatGateway1EIP.AllocationId
      SubnetId: !Ref PublicSubnet1
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-ngw-1

  NatGateway2:
    Type: AWS::EC2::NatGateway
    Properties:
      AllocationId: !GetAtt NatGateway2EIP.AllocationId
      SubnetId: !Ref PublicSubnet2
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-ngw-2

  NatGateway3:
    Type: AWS::EC2::NatGateway
    Properties:
      AllocationId: !GetAtt NatGateway3EIP.AllocationId
      SubnetId: !Ref PublicSubnet3
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-ngw-3

  PublicRouteTable:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VPC
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-public-routetable

  DefaultPublicRoute:
    Type: AWS::EC2::Route
    DependsOn: InternetGatewayAttachment
    Properties:
      RouteTableId: !Ref PublicRouteTable
      DestinationCidrBlock: 0.0.0.0/0
      GatewayId: !Ref InternetGateway

  PublicSubnet1RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PublicRouteTable
      SubnetId: !Ref PublicSubnet1

  PublicSubnet2RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PublicRouteTable
      SubnetId: !Ref PublicSubnet2

  PublicSubnet3RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PublicRouteTable
      SubnetId: !Ref PublicSubnet3

  PrivateRouteTable1:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VPC
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-routes-1

  DefaultPrivateRoute1:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable1
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway1

  PrivateSubnet1RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable1
      SubnetId: !Ref PrivateSubnet1

  PrivateRouteTable2:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VPC
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-routes-2

  DefaultPrivateRoute2:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable2
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway2

  PrivateSubnet2RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable2
      SubnetId: !Ref PrivateSubnet2

  PrivateRouteTable3:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VPC
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-routes-3

  DefaultPrivateRoute3:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable3
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway3

  PrivateSubnet3RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable3
      SubnetId: !Ref PrivateSubnet3

  NoIngressSecurityGroup:
    Type: AWS::EC2::SecurityGroup
    Properties:
      GroupName: "no-ingress-sg"
      GroupDescription: "Security group with no ingress rule"
      VpcId: !Ref VPC

Outputs:
  VPC:
    Description: A reference to the created VPC
    Value: !Ref VPC

  PublicSubnets:
    Description: A list of the public subnets
    Value: !Join [ ",", [ !Ref PublicSubnet1, !Ref PublicSubnet2 , !Ref PublicSubnet3 ]]

  PrivateSubnets:
    Description: A list of the private subnets
    Value: !Join [ ",", [ !Ref PrivateSubnet1, !Ref PrivateSubnet2 , !Ref PrivateSubnet3 ]]

  PublicSubnet1:
    Description: A reference to the public subnet in the 1st Availability Zone
    Value: !Ref PublicSubnet1

  PublicSubnet2:
    Description: A reference to the public subnet in the 2nd Availability Zone
    Value: !Ref PublicSubnet2

  PublicSubnet3:
    Description: A reference to the public subnet in the 3rd Availability Zone
    Value: !Ref PublicSubnet3

  PrivateSubnet1:
    Description: A reference to the private subnet in the 1st Availability Zone
    Value: !Ref PrivateSubnet1

  PrivateSubnet2:
    Description: A reference to the private subnet in the 2nd Availability Zone
    Value: !Ref PrivateSubnet2

  PrivateSubnet3:
    Description: A reference to the private subnet in the 3rd Availability Zone
    Value: !Ref PrivateSubnet3

  NatGateway1:
    Description: A reference to the Nat Gateway in the 1st Availability Zone
    Value: !Ref NatGateway1

  NatGateway2:
    Description: A reference to the Nat Gateway in the 2nd Availability Zone
    Value: !Ref NatGateway2

  NatGateway3:
    Description: A reference to the Nat Gateway in the 3rd Availability Zone
    Value: !Ref NatGateway3

  NoIngressSecurityGroup:
    Description: Security group with no ingress rule
    Value: !Ref NoIngressSecurityGroup
aws cloudformation create-stack --stack-name $CLUSTER_NAME-vpc \
      --template-body file://eks-vpc.yaml \
      --parameters \
ParameterKey=EnvironmentName,ParameterValue=$CLUSTER_NAME

Step 2:  You can create the Amazon EKS control plane in a number of ways by using tools like eksctl, cdk8s, or custom CloudFormation templates. For this example, we have gone ahead and used this CloudFormation stack by running the below AWS CLI command, which deploys the cluster on the subnets created in Step 1.

Save the below CloudFormation as eks-control-plane.yaml

AWSTemplateFormatVersion: '2010-09-09'

Description: Stack for creating an basic EKS control plane.

Parameters:
  Name:
    Type: String
    Description: EKS cluster name.
  Vpc:
    Type: AWS::EC2::VPC::Id
    Description: VPC to create the control plane and SGs in.
  Subnets:
    Type: List<AWS::EC2::Subnet::Id>
    Description: Subnets to create the control plane in. Recommend picking both public
      and private subnets.
  Version:
    Type: String
    Description: Kubernetes master version.
    AllowedValues:
      - '1.16'
      - '1.15'
      - '1.14'
    Default: '1.16'
    ConstraintDescription: Pick a version for the K8s master.

Resources:
  ClusterRole:
    Type: AWS::IAM::Role
    Properties:
      RoleName: !Join
        - '-'
        - - !Ref 'AWS::StackName'
          - cluster-role
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/AmazonEKSClusterPolicy
        - arn:aws:iam::aws:policy/AmazonEKSServicePolicy
        - arn:aws:iam::aws:policy/AmazonEC2ReadOnlyAccess
      AssumeRolePolicyDocument:
        Version: '2012-10-17'
        Statement:
          - Effect: Allow
            Principal:
              Service: eks.amazonaws.com
            Action: sts:AssumeRole
  ControlPlaneSecurityGroup:
    Type: AWS::EC2::SecurityGroup
    Properties:
      GroupDescription: Cluster communication with worker nodes
      VpcId: !Ref 'Vpc'
  ControlPlane:
    Type: AWS::EKS::Cluster
    Properties:
      Name: !Ref 'Name'
      ResourcesVpcConfig:
        SecurityGroupIds:
          - !Ref 'ControlPlaneSecurityGroup'
        SubnetIds: !Ref 'Subnets'
      RoleArn: !GetAtt 'ClusterRole.Arn'
      Version: !Ref 'Version'
Outputs:
  ClusterArn:
    Value: !GetAtt 'ControlPlane.Arn'
  ApiEndpoint:
    Value: !GetAtt 'ControlPlane.Endpoint'
  CertificateAuthorityData:
    Value: !GetAtt 'ControlPlane.CertificateAuthorityData'

Execute the below commands to run the cloudformation stack.

export VPC_ID=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc --query 'Stacks[0].Outputs[?OutputKey==`VPC`].OutputValue' --output text)

export SUBNETS_IDS=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc --query 'Stacks[0].Outputs[?OutputKey==`PrivateSubnets`].OutputValue' --output text)

aws cloudformation create-stack --stack-name $CLUSTER_NAME-control-plane \
    --template-body file://eks-control-plane.yaml \
    --capabilities CAPABILITY_NAMED_IAM \
    --parameters \
    ParameterKey=Name,ParameterValue=$CLUSTER_NAME \
    ParameterKey=Vpc,ParameterValue=$VPC_ID \    
'ParameterKey=Subnets,ParameterValue="'"$SUBNETS_IDS"'"'

Step 3: you will create the kubeconfig file for your cluster using the below command.  By default, the resulting configuration file is created at the default kubeconfig path (.kube/config) in your home directory or merged with an existing kubeconfig at that location.

aws eks update-kubeconfig --name $CLUSTER_NAME

At this point, you can verify if your service is up using the command “kubectl get svc” as shown below.

Figure 3 - kubernetes service

Step 4: we will use the CloudFormation stack to create the data plane, which is the first set of managed worker node groups, using the below AWS CLI command to deploy the CloudFormation stack. In this command, we feed in some of the inputs such as:

  • ClusterName, which is name of the cluster you used in step 1
  • A node group name
  • Node group auto scaling parameters like MinSize, MaxSize, and Desired Size
  • The key pair name to be used for the worker nodes (EC2 instances)
  • The security groups to protect the worker nodes
  • Subnet id’s where the worker nodes should be deployed and the VPC id

Save the below CloudFormation stack as eks-data-plane.yaml

AWSTemplateFormatVersion: '2010-09-09'

Description: Stack for creating worker nodes for an existing EKS cluster.

Parameters:
  ProvidedSecurityGroup:
    Type: String
    Description: Inbound SSH, Outbout ALL
    Default: 'sg-0d640ec8e209ef5a7'
  KeyName:
    Description: The EC2 Key Pair to allow SSH access to the instances
    Type: AWS::EC2::KeyPair::KeyName
  NodeInstanceType:
    Description: EC2 instance type for the node instances
    Type: String
    Default: t3.small
    ConstraintDescription: Must be a valid EC2 instance type
  AmiType:
    Type: String
    Description: The AMI type for the Node Group
    Default: AL2_x86_64
  NodeGroupName:
    Type: String
    Description: Unique identifier for the Node Group.
  NodeGroupScalingConfigMinSize:
    Type: Number
    Description: Minimum size of Node Group ASG
    Default: 3
  NodeGroupScalingConfigMaxSize:
    Type: Number
    Description: Maximum size of Node Group ASG. Set to at least 1 greater than NodeGroupScalingConfigDesiredSize
    Default: 10
  NodeGroupScalingConfigDesiredSize:
    Type: Number
    Description: Desired capacity of Node Group ASG
    Default: 3
  NodeDiskSize:
    Type: Number
    Description: Node EBS volume size
    Default: 300
  ClusterName:
    Description: The cluster name provided when the cluster was created. If it is
      incorrect, nodes will not be able to join the cluster
    Type: String
  ClusterControlPlaneSecurityGroup:
    Description: The security group of the cluster control plane
    Type: AWS::EC2::SecurityGroup::Id
  VpcId:
    Description: The VPC of the worker instances
    Type: AWS::EC2::VPC::Id
  Subnets:
    Description: The subnets where workers should be created. Recommended to pick
      only private subnets.
    Type: List<AWS::EC2::Subnet::Id>

Resources:
  NodeInstanceRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument:
        Version: '2012-10-17'
        Statement:
          - Effect: Allow
            Principal:
              Service:
                - ec2.amazonaws.com
            Action:
              - sts:AssumeRole
      Path: /
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/AmazonEKSWorkerNodePolicy
        - arn:aws:iam::aws:policy/AmazonEKS_CNI_Policy
        - arn:aws:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly
        - arn:aws:iam::aws:policy/AdministratorAccess
  NodeInstanceProfile:
    Type: AWS::IAM::InstanceProfile
    Properties:
      Path: /
      Roles:
        - !Ref 'NodeInstanceRole'
  NodeSecurityGroup:
    Type: AWS::EC2::SecurityGroup
    Properties:
      GroupDescription: Security group for all nodes in the cluster
      VpcId: !Ref 'VpcId'
      Tags:
        - Key: !Sub 'kubernetes.io/cluster/${ClusterName}'
          Value: owned
  NodeSecurityGroupIngress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow node to communicate with each other
      GroupId: !Ref 'NodeSecurityGroup'
      SourceSecurityGroupId: !Ref 'NodeSecurityGroup'
      IpProtocol: '-1'
      FromPort: 0
      ToPort: 65535
  NodeSecurityGroupKubernetesService:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow all traffic for K8s service ports. Workaround for NLB bug
      GroupId: !Ref 'NodeSecurityGroup'
      CidrIp: '0.0.0.0/0'
      IpProtocol: tcp
      FromPort: 30000
      ToPort: 32767
  NodeSecurityGroupFromControlPlaneIngress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow worker Kubelets and pods to receive communication from the
        cluster control plane
      GroupId: !Ref 'NodeSecurityGroup'
      SourceSecurityGroupId: !Ref 'ClusterControlPlaneSecurityGroup'
      IpProtocol: tcp
      FromPort: 1025
      ToPort: 65535
  ControlPlaneEgressToNodeSecurityGroup:
    Type: AWS::EC2::SecurityGroupEgress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow the cluster control plane to communicate with worker Kubelet
        and pods
      GroupId: !Ref 'ClusterControlPlaneSecurityGroup'
      DestinationSecurityGroupId: !Ref 'NodeSecurityGroup'
      IpProtocol: tcp
      FromPort: 1025
      ToPort: 65535
  NodeSecurityGroupFromControlPlaneOn443Ingress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow pods running extension API servers on port 443 to receive
        communication from cluster control plane
      GroupId: !Ref 'NodeSecurityGroup'
      SourceSecurityGroupId: !Ref 'ClusterControlPlaneSecurityGroup'
      IpProtocol: tcp
      FromPort: 443
      ToPort: 443
  ControlPlaneEgressToNodeSecurityGroupOn443:
    Type: AWS::EC2::SecurityGroupEgress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow the cluster control plane to communicate with pods running
        extension API servers on port 443
      GroupId: !Ref 'ClusterControlPlaneSecurityGroup'
      DestinationSecurityGroupId: !Ref 'NodeSecurityGroup'
      IpProtocol: tcp
      FromPort: 443
      ToPort: 443
  ClusterControlPlaneSecurityGroupIngress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow pods to communicate with the cluster API Server
      GroupId: !Ref 'ClusterControlPlaneSecurityGroup'
      SourceSecurityGroupId: !Ref 'NodeSecurityGroup'
      IpProtocol: tcp
      ToPort: 443
      FromPort: 443
  NodeSSHSecurityGroupIngress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow ssh into worker node
      GroupId: !Ref 'NodeSecurityGroup'
      SourceSecurityGroupId: !Ref 'ProvidedSecurityGroup'
      IpProtocol: tcp
      ToPort: 22
      FromPort: 22
  NodeGroup:
    Type: AWS::EKS::Nodegroup
    Properties:
      AmiType: !Ref 'AmiType'
      ClusterName: !Ref 'ClusterName'
      DiskSize: !Ref 'NodeDiskSize'
      InstanceTypes:
        - !Ref 'NodeInstanceType'
      NodegroupName: !Ref 'NodeGroupName'
      NodeRole: {"Fn::GetAtt" : ["NodeInstanceRole", "Arn"] }
      RemoteAccess:
        Ec2SshKey: !Ref 'KeyName'
        SourceSecurityGroups:
          - !Ref 'ProvidedSecurityGroup'
          - !Ref 'NodeSecurityGroup'
      ScalingConfig:
        MinSize: !Ref 'NodeGroupScalingConfigMinSize'
        DesiredSize: !Ref 'NodeGroupScalingConfigDesiredSize'
        MaxSize: !Ref 'NodeGroupScalingConfigMaxSize'
      Subnets: !Ref 'Subnets'
      Tags:
        Name: !Sub ${NodeGroupName}-NodeGroup
Outputs:
  NodeInstanceRole:
    Description: The node instance role
    Value: !GetAtt 'NodeInstanceRole.Arn'
  NodeSecurityGroup:
    Description: The security group for the node group
    Value: !Ref 'NodeSecurityGroup'
 
       
export CLUSTER_SG=$(aws eks describe-cluster --name $CLUSTER_NAME --query 'cluster.resourcesVpcConfig.clusterSecurityGroupId')

export ADDITIONAL_SG=$(aws eks describe-cluster --name $CLUSTER_NAME --query 'cluster.resourcesVpcConfig.securityGroupIds[0]')

aws cloudformation create-stack --stack-name $CLUSTER_NAME-data-plane \
--template-body file://eks-data-plane.yaml \
--capabilities CAPABILITY_NAMED_IAM \
--parameters \
ParameterKey=ClusterControlPlaneSecurityGroup,ParameterValue=$CLUSTER_SG \
ParameterKey=ClusterName,ParameterValue=$CLUSTER_NAME \
ParameterKey=NodeGroupName,ParameterValue=$CLUSTER_NAME-nodegroup \
ParameterKey=KeyName,ParameterValue=$KEY_NAME \
ParameterKey=VpcId,ParameterValue=$VPC_ID \
'ParameterKey=Subnets,ParameterValue="'"$SUBNETS_IDS"'"' \
ParameterKey=ProvidedSecurityGroup,ParameterValue=$ADDITIONAL_SG
 
       

Before deploying the application, you can use the below command to monitor and wait until all the nodes are in “Ready” state.

kubectl get nodes --watch

Step 5: We will use this deployment.yaml file to deploy a demo application using the below kubectl command. Save the below file as deployment.yaml file

apiVersion: apps/v1 
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 2 # tells deployment to run 2 pods matching the template
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginxdemos/hello:latest
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: nginx
  namespace: default
  labels:
    app: nginx
spec:
  ports:
  - name: http
    port: 80
    protocol: TCP
    targetPort: 80
  selector:
    app: nginx
  type: LoadBalancer

Execute the below command to deploy the app.

kubectl create -f deployment.yaml

This will take a few minutes after which the pods should be in running state. You can verify using the below command:

kubectl get Pods --all-namespaces

The below command will display the load balancer endpoint of the deployed application for testing:

kubectl get service -o wide

Figure 4 - Kubernetes service

Now, when you open this endpoint on a browser, you will observe that the server is deployed on the 10.0.0.0/16 CIDR block.

Figure 5 - Application accessed on browser

Step 6: In this step, we will attach a new CIDR block to the VPC and create new subnets on the 100.64.0.0/16 CIDR block using this CloudFormation stack and the below AWS CLI commands.

You may save the below CloudFormation as eks-vpc-secondary.yaml

AWSTemplateFormatVersion: '2010-09-09'

Description:  This template deploys a new CIDR block and three private subnets spread
  across three Availability Zones. 

Parameters:
  EnvironmentName:
    Description: An environment name that is prefixed to resource names
    Type: String
  VpcId:
    Description: VPC ID of the main VPC
    Type: String
  VpcCIDR:
    Description: Please enter the IP range (CIDR notation) for this VPC
    Type: String
    Default: 100.64.0.0/16
  PrivateSubnet1CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the first Availability Zone
    Type: String
    Default: 100.64.0.0/20
  PrivateSubnet2CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the second Availability Zone
    Type: String
    Default: 100.64.16.0/20
  PrivateSubnet3CIDR:
    Description: Please enter the IP range (CIDR notation) for the private subnet in the second Availability Zone
    Type: String
    Default: 100.64.32.0/20
  NatGateway1:
    Description: Please enter the Nat Gateway in the first Availability Zone
    Type: String
  NatGateway2:
    Description: Please enter the Nat Gateway in the second Availability Zone
    Type: String
  NatGateway3:
    Description: Please enter the Nat Gateway in the third Availability Zone
    Type: String


Resources:
  VpcCidrBlock:
    Type: AWS::EC2::VPCCidrBlock
    Properties:
      VpcId: !Ref VpcId
      CidrBlock: !Ref VpcCIDR

  PrivateSubnet1:
    Type: AWS::EC2::Subnet
    DependsOn: VpcCidrBlock
    Properties:
      VpcId: !Ref VpcId
      AvailabilityZone: !Select [ 0, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet1CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-subnet-1
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateSubnet2:
    Type: AWS::EC2::Subnet
    DependsOn: VpcCidrBlock
    Properties:
      VpcId: !Ref VpcId
      AvailabilityZone: !Select [ 1, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet2CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-subnet-2
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateSubnet3:
    Type: AWS::EC2::Subnet
    DependsOn: VpcCidrBlock
    Properties:
      VpcId: !Ref VpcId
      AvailabilityZone: !Select [ 2, !GetAZs  '' ]
      CidrBlock: !Ref PrivateSubnet3CIDR
      MapPublicIpOnLaunch: false
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-subnet-3
        - Key:  !Sub kubernetes.io/cluster/${EnvironmentName}
          Value: shared

  PrivateRouteTable1:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VpcId
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-routes-1

  DefaultPrivateRoute1:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable1
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway1

  PrivateSubnet1RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable1
      SubnetId: !Ref PrivateSubnet1

  PrivateRouteTable2:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VpcId
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-routes-2

  DefaultPrivateRoute2:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable2
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway2

  PrivateSubnet2RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable2
      SubnetId: !Ref PrivateSubnet2

  PrivateRouteTable3:
    Type: AWS::EC2::RouteTable
    Properties:
      VpcId: !Ref VpcId
      Tags:
        - Key: Name
          Value: !Sub ${EnvironmentName}-private-secondary-routes-3

  DefaultPrivateRoute3:
    Type: AWS::EC2::Route
    Properties:
      RouteTableId: !Ref PrivateRouteTable3
      DestinationCidrBlock: 0.0.0.0/0
      NatGatewayId: !Ref NatGateway3

  PrivateSubnet3RouteTableAssociation:
    Type: AWS::EC2::SubnetRouteTableAssociation
    Properties:
      RouteTableId: !Ref PrivateRouteTable3
      SubnetId: !Ref PrivateSubnet3

Outputs:

  PrivateSubnets:
    Description: A list of the private subnets
    Value: !Join [ ",", [ !Ref PrivateSubnet1, !Ref PrivateSubnet2 , !Ref PrivateSubnet3 ]]

  PrivateSubnet1:
    Description: A reference to the private subnet in the 1st Availability Zone
    Value: !Ref PrivateSubnet1

  PrivateSubnet2:
    Description: A reference to the private subnet in the 2nd Availability Zone
    Value: !Ref PrivateSubnet2

  PrivateSubnet3:
    Description: A reference to the private subnet in the 3rd Availability Zone
    Value: !Ref PrivateSubnet3

Run the following commands to deploy the stack.

export NGW1=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc --query 'Stacks[0].Outputs[?OutputKey==`NatGateway1`].OutputValue' --output text) 

export NGW2=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc --query 'Stacks[0].Outputs[?OutputKey==`NatGateway2`].OutputValue' --output text) export NGW3=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc --query 'Stacks[0].Outputs[?OutputKey==`NatGateway3`].OutputValue' --output text) 

aws cloudformation create-stack --stack-name $CLUSTER_NAME-vpc-secondary \
 --template-body file://eks-vpc-secondary.yaml \
 --parameters \ ParameterKey=EnvironmentName,ParameterValue=$CLUSTER_NAME \
 ParameterKey=VpcId,ParameterValue=$VPC_ID \
 ParameterKey=NatGateway1,ParameterValue=$NGW1 \
 ParameterKey=NatGateway2,ParameterValue=$NGW2 \
 ParameterKey=NatGateway3,ParameterValue=$NGW3

Step 7:  In this step, we will deploy the second set of managed worker nodes on the new subnets we created in the previous step using this CloudFormation and the below AWS CLI command:

You will be using the CloudFormation stack created in step 4 (eks-data-plane.yaml).

export SECONDARY_SUBNETS_IDS=$(aws cloudformation describe-stacks --stack-name $CLUSTER_NAME-vpc-secondary --query 'Stacks[0].Outputs[?OutputKey==`PrivateSubnets`].OutputValue' --output text)

aws cloudformation create-stack --stack-name $CLUSTER_NAME-data-plane-secondary \
--template-body file://eks-data-plane.yaml \
--capabilities CAPABILITY_NAMED_IAM \
--parameters \
ParameterKey=ClusterControlPlaneSecurityGroup,ParameterValue=$CLUSTER_SG \
ParameterKey=ClusterName,ParameterValue=$CLUSTER_NAME \
ParameterKey=NodeGroupName,ParameterValue=$CLUSTER_NAME-nodegroup-secondary \
ParameterKey=KeyName,ParameterValue=$KEY_NAME \
ParameterKey=VpcId,ParameterValue=$VPC_ID \
'ParameterKey=Subnets,ParameterValue="'"$SECONDARY_SUBNETS_IDS"'"' \
ParameterKey=ProvidedSecurityGroup,ParameterValue=$ADDITIONAL_SG
 
       

You can wait until the nodes are in “Ready” state using the below command:

kubectl get nodes --watch

Step 8: Now that we have the application running on subnets using 10.0.0.0/8 CIDR block, we will first issue the cordon command. This command stops scheduling any new pods to these worker nodes. We will then drain these worker nodes deployed in step 4, so that the pods automatically get terminated and recreated in other healthy worker nodes based on subnets using 100.64.0.0/16 CIDR range.

Command to cordon all the nodes running on 10.0.0.0/16 CIDR block.

kubectl get nodes --no-headers=true | awk '/ip-10-0/{print $1}' | xargs kubectl cordon

Command to drain all the nodes running on 10.0.0.0/16 CIDR block. You need to use –ignore-daemonsets flag in order to drain nodes with daemonsets and use –delete-local-data flag to overide and delete any pods that utilize an emptyDir volume.

kubectl get nodes --no-headers=true | awk '/ip-10-0/{print $1}' | xargs kubectl drain --force --ignore-daemonsets --delete-local-data

You can verify that the endpoint is now running on 100.64.0.0/16 CIDR block by accessing the endpoint again using the below command:

kubectl get service -o wide

Conclusion

You should now be able to choose the most appropriate solution if you are running into an issue of limited IP addresses in your existing VPC CIDR range. Also, it is very important to plan out your VPC CIDR ranges across your multiple accounts to make sure you do not have overlaping IP addresses requiring complex NATing resolutions. We highly recommend reading this blog that explains various networking patterns, especially for customers having hybrid environments with connectivity to their data centers.

Further reading references:

If you are new to Kubernetes, this should help you understand the basics. Also, you can check our documentation if you are new to Amazon EKS. With the adoption of Kubernetes, Jeremy Cowan explains in this blog on how the problem IP address shortage escalates with increase in workload and also provides a solution to this problem.

 

About the Authors:

Jose Olcese

Jose is a Principal Cloud Application Architect with Amazon Web Services where he helps customers build cutting-edge, cloud-based solutions. Jose has over 20 years of experience in software development for a variety of different industries and has helped hundreds of customers to integrate Identity solutions with AWS. Outside of work, Jose enjoys spending time with his family, running and building things.

 

Umesh Kumar Ramesh

Umesh is a Sr. Cloud Infrastructure Architect with AWS who delivers proof-of-concept projects, topical workshops, and leads implementation projects. He holds a Bachelor’s degree in Computer Science & Engineering from the National Institute of Technology, Jamshedpur (India). Outside of work, he enjoys watching documentaries, biking, practicing meditation and discuss spirituality.

TAGS:
Umesh Ramesh

Umesh Ramesh