AKS Cluster with a NAT Gateway and an Application Gateway

Azure Public Test Date Azure Public Test Result

Azure US Gov Last Test Date Azure US Gov Last Test Result

Best Practice Check Cred Scan Check

Bicep Version

Deploy To Azure

Visualize

This ARM template can be used to deploy a public or private Azure Kubernetes Cluster (AKS) cluster with an Azure Application Gateway and Application Gateway Ingress Controller add-on. Azure CNI and Dynamic allocation of IPs and enhanced subnet support are used to assign a private IP address to each pod from a subnet separate from the subnet hosting the AKS cluster. In addition, this ARM template shows how to manage outbound connections generated by workloads running on AKS using a Azure NAT Gateway and an Azure Public IP address prefix. Virtual Network NAT or NAT Gateway is a fully managed and highly resilient Network Address Translation (NAT) service. VNet NAT simplifies outbound Internet connectivity for virtual networks. One more Public IP Addresses or Public IP Prefix can be used by a single Azure NAT Gateway. When configured on a subnet, all outbound connectivity uses the VNet NAT's static public IP addresses. A public IP address prefix is a reserved range of public IP addresses in Azure. Public IP prefixes are assigned from a pool of addresses in each Azure region. You create a public IP address prefix in an Azure region and subscription by specifying a name and prefix size. The prefix size is the number of addresses available for use. Public IP address prefixes consist of IPv4 or IPv6 addresses. In regions with Availability Zones, Public IP address prefixes can be created as zone-redundant or associated with a specific availability zone. After the public IP prefix is created, you can create public IP addresses. Azure NAT Gateway allows up to 64,000 outbound UDP and TCP traffic flows per IP address with a maximum of 16 IP addresses. When you create a new AKS cluster, you can specify how oubound connections from the agent nodes will be handled by specifying one of the following values for the outboundType property:

  • loadBalancer: outbound connections and SNAT ports are handled by the outbound rules of os standard load balancer used by the AKS cluster. For more information, see Outbound rules Azure Load Balancer.
  • managedNATGateway: a managed NAT Gateway is provisioned by the Microsoft.ContainerService resource provider in the node resource group and associated to the subnets hosting the node pools. For more information, see Managed NAT Gateway.
  • userAssignedNATGateway: the AKS cluster is configured to use a user-defined NAT Gateway. In this case, it's your responsibility to create the NAT Gateway and associate it to the subnets hosting the node pools of the AKS cluster.
  • userDefinedRouting: the outbound connections from the AKS cluster are routed to a firewall like Azure Firewall located in the same virtual network or in a perimiter network peered to the AKS cluster virtual network via a route table and user-defined routes. For more information, see Customize cluster egress with a User-Defined Route and Create a private Azure Kubernetes Service cluster behind Azure Firewall using Terraform and Azure DevOps.

The template allows to deploy a rich set of AKS features such as:

For a sample on how to use the Application Gateway Ingress Controller in a multi-tenant AKS cluster, see this project.

Architecture

This diagram shows the overall architecture:

Architecture

The ARM template deploys:

  • A new virtual network with five subnets:
    • AksSubnet: this subnet is used for the AKS cluster worker nodes. The VMSS of both the system and user node pools will be created in this subnet.
    • PodSubnet: this subnet is used to allot private IP addresses to pods of both the system and user node pools.
    • VmSubnet: hosts the jumpbox virtual machine and private endpoints
    • AppGatewaySubnet: hosts Application Gateway WAF2
    • AzureBastionSubnet: Azure Bastion
  • The AKS cluster uses a user-defined managed identity to create additional resources like load balancers and managed disks in Azure.
  • The AKS cluster is composed of a:
    • System node pool hosting only critical system pods and services. The worker nodes have node taint which prevents application pods from beings scheduled on this node pool.
    • User node pool hosting user workloads and artifacts.
  • A NAT Gateway used to handle outbound connections from the AksSubnet, PodSubnet, and VmSubnet.
  • An Azure Bastion resource that provides secure and seamless SSH connectivity to the jumpbox virtual machine directly in the Azure portal over SSL
  • An Azure Container Registry (ACR) to build, store, and manage container images and artifacts in a private registry for all types of container deployments.
  • a Storage Account for the boot diagnostic logs of the jumpbox virtual machine
  • An Application Gateway used by the Application Gateway Ingress Controller
  • A Web Access Firewall (WAF) Policy associated to the Application Gateway as the root level and HTTP listener level. The Policy is configured in Prevention mode and uses the OWASP 3.1 rule set and a couple of custom rules that demostrate how to block requests when the query string or a header contain a specific string. You can create more sophisticated custom rules to whitelist or blacklist the incoming traffic to the Application Gateway Ingres Controller.
  • A Key Vault that can be used by workloads running on AKS to retrieve keys, certificates, and secrets
  • A private endpoint to the Blob Storage Account
  • A private endpoint to to Azure Container Registry (ACR)
  • A private endpoint to Key Vault
  • When you opt for a private AKS cluster, a private endpoint to the control plane / API server hosted by an AKS-managed Azure subscription. In this case, the cluster can communicate with the API server exposed via a Private Link Service using a private endpoint.
  • When you choose Premium as SKU for ACR, a Private Endpoint is created to allow the private AKS cluster to access ACR via a private IP address. For more information, see Connect privately to an Azure container registry using Azure Private Link.
  • A Private DNS Zone for the name resolution of the private endpoint to the Blob Storage Account
  • A Private DNS Zone for the name resolution of the private endpoint to Azure Container Registry (ACR)
  • A Private DNS Zone for the name resolution of the private endpoint to Key Vault
  • When you deploy the cluster as private, a Private DNS Zone for the name resolution of the private endpoint to the Kubernetes Server API
  • A Virtual Network Link between the virtual network hosting the AKS cluster and the above Private DNS Zones
  • A jumpbox virtual machine to manage the AKS cluster in case you decide to deploy a private AKS cluster.
  • A Log Analytics workspace to collect the diagnostics logs and metrics from:
    • AKS cluster
    • Jumpbox virtual machine
    • Application Gateway
    • Key Vault
    • Network Security Group

Application Gateway Ingress Controller: Deployment Options

In this architecture, the Application Gateway Ingress Controller was installed using the AGIC add-on for AKS. You can also install the Application Gateway Ingress Controller via a Helm chart. The primary benefit of deploying AGIC as an AKS add-on is that it's much simpler than deploying through Helm. For a new setup, you can deploy a new Application Gateway and a new AKS cluster with AGIC enabled as an add-on in one line in Azure CLI. The add-on is also a fully managed service, which provides added benefits such as automatic updates and increased support. Both ways of deploying AGIC (Helm and AKS add-on) are fully supported by Microsoft. Additionally, the add-on allows for better integration with AKS as a first class add-on.

The AGIC add-on is still deployed as a pod in the customer's AKS cluster, however, there are a few differences between the Helm deployment version and the add-on version of AGIC. Below is a list of differences between the two versions:

  • Helm deployment values cannot be modified on the AKS add-on:

    • verbosityLevel will be set to 5 by default
    • usePrivateIp will be set to be false by default; this can be overwritten by the use-private-ip annotation
    • shared is not supported on add-on
    • reconcilePeriodSeconds is not supported on add-on
    • armAuth.type is not supported on add-on
  • AGIC deployed via Helm supports ProhibitedTargets, which means AGIC can configure the Application Gateway specifically for AKS clusters without affecting other existing backends. AGIC add-on doesn't currently support this.

  • Since AGIC add-on is a managed service, customers will automatically be updated to the latest version of AGIC add-on, unlike AGIC deployed through Helm where the customer must manually update AGIC.

  • As shown in the following picture, customers can only deploy one AGIC add-on per AKS cluster, and each AGIC add-on currently can only target one Application Gateway. For deployments that require more than one AGIC per cluster or multiple AGICs targeting one Application Gateway, please continue to use AGIC deployed through Helm.

setup

As documented at Enable multiple Namespace support in an AKS cluster with Application Gateway Ingress Controller, a single instance of the Azure Application Gateway Kubernetes Ingress Controller (AGIC) can ingest events from and observe multiple namespaces. Should the AKS administrator decide to use App Gateway as an ingress, all namespaces will use the same instance of Application Gateway. A single installation of Ingress Controller will monitor accessible namespaces and will configure the Application Gateway it is associated with.

To enable multiple namespace support:

  • modify the helm-config.yaml file in one of the following ways:

    • delete the watchNamespace key entirely from helm-config.yaml - AGIC will observe all namespaces
    • set watchNamespace to an empty string - AGIC will observe all namespaces
    • add multiple namespaces separated by a comma (watchNamespace: default,secondNamespace) - AGIC will observe these namespaces exclusively
  • apply Helm template changes with: helm install -f helm-config.yaml application-gateway-kubernetes-ingress/ingress-azure

Once deployed with the ability to observe multiple namespaces, AGIC will:

  • list ingress resources from all accessible namespaces
  • filter to ingress resources annotated with kubernetes.io/ingress.class: azure/application-gateway
  • compose combined Application Gateway config
  • apply the config to the associated Application Gateway via ARM

Limits

Azure subscription and service limits, quotas, and constraints documentation reports that the max number of:

  • Active Listeners
  • Backend Pools
  • HTTP Load Balancing Rules
  • HTTP Settings
  • Authentication certificates
  • Etc.

is 100, and in case of WAF-enabled SKUs, this limit is 40. This implies that the maximum number of tenants that can be served by a single AGIC is equal to 100 when using Application Gateway Standard V2, and 40 for Application Gateway WAF V2.

Dynamic Public IP Allocation

A drawback with the traditional CNI is the exhaustion of pod IP addresses as the AKS cluster grows, resulting in the need to rebuild the entire cluster in a bigger subnet. The new Dynamic IP Allocation capability in Azure CNI solves this problem by allotting pod IP addresses from a subnet separate from the subnet hosting the AKS cluster nodes. This feature offers the following benefits:

  • Better IP utilization: private IP addresses are dynamically allocated to cluster Pods from the Pod subnet. This leads to better utilization of private IP addresses in the cluster compared to the traditional CNI solution, which statically allocate to each worker node the same number of private IP addresses from the subnet.
  • Scalable and flexible: when using separate subnets for nodes and pods, the two subnets can be scaled independently. A single pod subnet can be shared across multiple node pools of a cluster or across multiple AKS clusters deployed in the same virtual network. You can also configure a separate pod subnet for a node pool. Likewise, you can deploy node pools in the same subnet or in separate node pools. You can define the subnet for worker nodes and pods of a node pool at provisioning time.
  • High performance: Since pod are assigned private IP addresses from a subnet, they have direct connectivity to other cluster pod and resources in the same virtual network or any peered virtual network. The solution supports very large clusters without any degradation in performance.
  • Separate VNet policies for pods: Since pods have a separate subnet, you can configure separate virtual network policies for them that are different from node policies. This enables many useful scenarios such as allowing internet connectivity only for pods and not for nodes, fixing the source IP for pod in a node pool using a NAT Gateway associated to the subnet hosting pods, and using NSGs to filter traffic between node pools.

Deployment

You can use the deploy.sh Bash script to deploy the topology. The following picture shows the resources deployed by the ARM template in the target resource group.

Resource Group

The following picture shows the resources deployed by the ARM template in the MC resource group associated to the AKS cluster:

MC Resource Group

NOTE: if you deploy the ARM template without using the companion deploy.sh, make sure to properly install the necessary preview features (for more information, see the script) and specify a value fo the following parameters:

  • aksClusterKubernetesVersion: The latest version of Kubernetes available in your region
  • aadProfileAdminGroupObjectIDs: an array containing the objectId of one or more Azure Active Directory groups for cluster administrator users
  • vmAdminPasswordOrKey: the key or password for the jumpbox virtual machine

Application Gateway

Azure Application Gateway is a web traffic load balancer that enables customers to manage the inbound traffic to multiple downstream web applications and REST APIs. Traditional load balancers operate at the transport layer (OSI layer 4 - TCP and UDP) and route traffic based on source IP address and port, to a destination IP address and port. The Application Gateway instead is an application layer (OSI layer 7) load balancer. Azure Application Gateway provides a rich set of features:

For more information, see How an Application Gateway works.

Web Access Firewall Policy for Application Gateway

Web Application Firewall (WAF) is a service that provides centralized protection of web applications from common exploits and vulnerabilities. WAF is based on rules from the OWASP (Open Web Application Security Project) core rule sets. WAF also provide the ability to create custom rules that are evaluated for each request that passes through the WAF. These rules hold a higher priority than the rest of the rules in the managed rule sets. The custom rules contain a rule name, rule priority, and an array of matching conditions. If these conditions are met, an action is taken (to allow or block). Web applications can be the target of malicious attacks that exploit common, known vulnerabilities that include SQL injection attacks, DDOS attacks, and cross site scripting attacks. Preventing such attacks in application code can be challenging and may require rigorous maintenance, patching and monitoring at many layers of the application topology. A centralized web application firewall helps make security management much simpler and gives better assurance to application administrators against threats or intrusions. A WAF solution can also react to a security threat faster by patching a known vulnerability at a central location versus securing each of individual web applications. Existing application gateways can be converted to a Web Application Firewall enabled application gateway very easily. Azure Application Gateway allows the association of a separate WAF policy to each individual listener. For example, if there are three sites behind the same Application Gateway or WAF, you can configure three separate WAF policies (one for each listener) to customize the exclusions, custom rules, and managed rulesets for one site without effecting the other two. If you want a single policy to apply to all sites, you can just associate the policy with the Application Gateway, rather than the individual listeners, to make it apply globally. Application Gateway also supports per-URI WAF Policies. This feature requires the use of a Path-based routing rule instead of a basic routing rule and requires the definition of a URL Path Map where a specific WAF policy can be associated to a given URL. For more information, see Configure per-site WAF policies using Azure PowerShell. The order of precedence for WAF policies is as follows:

  • If a per-URI WAF policy exists for the current path, this will take effect / apply and no other WAF policy will apply
  • If no per-URI WAF policy exists for the current path, but a WAF policy exists for the current listener, this policy will apply, and no other WAF policy will take effect
  • If no WAF policy exists for the current URI and listener, the global WAF policy will apply, if any.

The Application Gateway WAF can be configured to run in the following two modes:

  • Detection mode: Monitors and logs all threat alerts. You turn on logging diagnostics for Application Gateway in the Diagnostics section. You must also make sure that the WAF log is selected and turned on. Web application firewall doesn't block incoming requests when it's operating in Detection mode.
  • Prevention mode: Blocks intrusions and attacks that the rules detect. The attacker receives a "403 unauthorized access" exception, and the connection is closed. Prevention mode records such attacks in the WAF logs.

You can configure Application Gateway to store diagnostic logs and metrics to Log Analytics. In this case, also WAF logs will be stored in Log Analytics and they can be queries using Kusto Query Language. In the ARM template the WAF policy is configured in Prevention mode and contains a couple of sample custom rules that block incoming request, when the query string contains the word blockme or when the User-Agent header contain the string evilbot:

{
  "name": "BlockMe",
  "priority": 1,
  "ruleType": "MatchRule",
  "action": "Block",
  "matchConditions": [
    {
      "matchVariables": [
        {
          "variableName": "QueryString"
        }
      ],
      "operator": "Contains",
      "negationConditon": false,
      "matchValues": [
        "blockme"
      ],
      "transforms": []
    }
  ]
},
{
  "name": "BlockEvilBot",
  "priority": 2,
  "ruleType": "MatchRule",
  "action": "Block",
  "matchConditions": [
    {
      "matchVariables": [
        {
          "variableName": "RequestHeaders",
          "selector": "User-Agent"
        }
      ],
      "operator": "Contains",
      "negationConditon": false,
      "matchValues": [
        "evilbot"
      ],
      "transforms": [
        "Lowercase"
      ]
    }
  ]
}

Visio

The Visio document contains the architecture diagram.

References

Azure Kubernetes Service:

Azure Application Gateway:

Azure Application Gateway Ingress Controller

Azure Application Gateway WAF

Tags: Microsoft.Network/publicIPPrefixes, Microsoft.Network/natGateways, Microsoft.Network/publicIPAddresses, Microsoft.Network/bastionHosts, providers/diagnosticSettings, Microsoft.Storage/storageAccounts, Microsoft.Network/networkInterfaces, Microsoft.Compute/virtualMachines, Microsoft.Compute/virtualMachines/extensions, OmsAgentForLinux, DependencyAgentLinux, Microsoft.Network/networkSecurityGroups, Microsoft.Network/virtualNetworks, Microsoft.ManagedIdentity/userAssignedIdentities, Microsoft.Authorization/roleAssignments, Microsoft.KeyVault/vaults, Microsoft.ContainerRegistry/registries, Notary, Microsoft.ContainerService/managedClusters, UserAssigned, [parameters('systemNodePoolType')], [parameters('userNodePoolType')], Microsoft.OperationalInsights/workspaces, Microsoft.OperationsManagement/solutions, Microsoft.Network/privateDnsZones, Microsoft.Network/privateDnsZones/virtualNetworkLinks, Microsoft.Network/privateEndpoints, privateDnsZoneGroups, microsoft.insights/activityLogAlerts, Microsoft.Network/ApplicationGatewayWebApplicationFirewallPolicies, Microsoft.Network/applicationGateways