Container Service for Kubernetes:Create and manage Windows node pools in ACK clusters

Console

You cannot create the default node pool (default-nodepool) as a Windows node pool. See the Limits section above and create a Windows node pool after creating a cluster.

1. Log on to the ACK console. In the navigation pane on the left, click Clusters.

2. On the Clusters page, find the cluster to manage and click its name. In the left-side navigation pane, choose Nodes > Node Pools.

3. In the upper-right corner of the Node Pools page, click Create Node Pool, configure the parameters, and then click Confirm Order.

   For more information about the parameters, see Create and manage node pools. The following table describes some of the parameters.

   Parameter	Description
   Managed Node Pool	Select not to enable this feature.
   Container Runtime	Select containerd.
   Instance Type	Some instance types do not support Windows node pools. When you create a Windows node pool, only the instance types that support Windows node pools can be selected. For example, you can select the ecs.g6.xlarge instance type when you create a Windows node pool. Select an instance type with at least 4 vCPUs and 8 GB of memory. Select non-GPU-accelerated instance types. For more information about instance types, see Instance families.
   Expected Number of Nodes	The number of nodes that you want the node pool to maintain. After you specify the desired size, the node pool automatically scales in or out to maintain the specified number of nodes.
   Operating System	Select the Windows operating system.
   ECS Tags	Add labels to the Elastic Compute Service (ECS) instances.
   Node Labels	The labels that you want to add to the nodes in the node pool.

   On the Node Pools page, when the Status of the node pool is Ready, the Windows node pool is created.

API

You can call the CreateClusterNodePool operation to create a Windows node pool and configure parameters such as the number of nodes and instance types. The following example shows the parameters:

{
  "ClusterId": "ca8ed83XXXXXXXXXXXX5ec197",          // The ID of the cluster.
  "body": { 
    "nodepool_info": {
      "name": "nodepool-fXXXXXx",                    // The name of the node pool.
      "resource_group_id": "rg-acfmXXXXXXXXsjpq"     // The ID of the resource group to which the cluster belongs. This parameter is used to isolate different resources.  
    },
    "scaling_group": {
      "vswitch_ids": [                               // The vSwitches where the nodes of the cluster are deployed. This parameter is required when you create a managed cluster without nodes. 
        "vsw-bp1XXXXXXXXXXXXX1xm84",
        "vsw-bp1XXXXXXXXXXXX2v6x0a",
        "vsw-bp1XXXXXXXXXXXXXjkxbo"
      ],
      "data_disks": [                                // The configurations of the data disks attached to the worker nodes, such as the disk type and disk size.
        {
          "category": "cloud_essd",
          "size": 120,
          "encrypted": "false",
          "performance_level": "PL0"
        }
      ],
      "vpc_id": "vpc-bpXXXXXXXXXXXXXXXmgb6",         // The ID of the VPC where the cluster is deployed.
      "instance_charge_type": "PostPaid",            // The billing method of the nodes. Valid values: PrePaid: subscription. PostPaid: pay-as-you-go. Default value: PostPaid.
      "internet_max_bandwidth_out": 1,               // The maximum outbound bandwidth of the public IP address of the node. Unit: Mbit/s. Valid values: 1 to 100.
      "soc_enabled": false,                          // Specifies whether to enable MLPS security hardening. Valid values: true: enables MLPS security hardening. false: disables MLPS security hardening.
      "security_hardening_os": false,                // Specifies whether to enable security hardening for the operating system. Valid values: true: enables security hardening. false: disables security hardening. Default value: false.
      "instance_types": [                            // The instance types.
        "ecs.n2.large"
      ],
      "instance_patterns": [                         // The instance configurations.
        {
          "Cores": 4,
          "Memory": 16,
          "burst_performance_option": "Exclude",     // Specifies whether the instance type is a burstable instance type.  
          "excluded_instance_types": [               // The instance types that you want to exclude.
            "ecs.c6t.*"
            // ...
          ],
          "instance_type_families": [                // The instance families that you want to specify.  
            "ecs.g6"
            // ...
          ]
        }
      ],
      "login_password": "XXXXXX",
      "login_as_non_root": false,
      "system_disk_category": "cloud_essd",
      "system_disk_size": 40,
      "system_disk_performance_level": "PL0",
      "system_disk_encrypted": false,
      "multi_az_policy": "BALANCE",                // The scaling policy for the multi-zone scaling group.
      "platform": "WindowsServer2022",
      "image_id": "win2022_21H2_x64_dtc_en-us_40G_container_alibase_20240220.vhd",
      "image_type": "WindowsServer2022",
      "desired_size": 1
    },
    "kubernetes_config": {                         // The configurations of the cluster.
      "cms_enabled": true,                         // Specifies whether to install the CloudMonitor agent on ECS nodes. After the CloudMonitor agent is installed, you can view monitoring information about the ECS instances in the CloudMonitor console. We recommend that you install the CloudMonitor agent.
      "unschedulable": false,                      // Specifies whether to set the nodes that are added to the node pool as unschedulable. Valid values: true: The nodes are unschedulable after a scale-out activity is performed. false: The nodes are schedulable after a scale-out activity is performed.
      "runtime": "containerd",                   
      "runtime_version": "2.1.1"                   // Only containerd 1.6 and later versions support Windows node pools.
    },
    "auto_scaling": {                              // The configurations of auto scaling for the node pool. Valid values: true: enables auto scaling for the node pool. false: disables auto scaling.
      "enable": false
    },
    "tee_config": {                                // The configurations of confidential computing for the node pool. Valid values: true: enables confidential computing for the cluster. false: does not enable confidential computing for the cluster.
      "tee_enable": false
    },
    "node_config": {                               // The configurations of the nodes.    
      "image_acceleration_config": {
        "enable_image_acceleration": false
      }
    }
  }
}

Terraform

You can refer to the Terraform overview document to create a Windows node pool and configure parameters such as the number of nodes and instance types. The following example shows the parameters:

provider "alicloud" {
  region = "cn-hangzhou"
}

resource "alicloud_cs_kubernetes_node_pool" "pool" {
  tee_config {
    tee_enable = false
  }

  runtime_version   = "2.1.1"                    // The container runtime version. Only containerd 1.6 and later versions support Windows node pools.
  resource_group_id = "rg-acXXXXXXXjpq"
  multi_az_policy   = "BALANCE"
  node_pool_name    = "nodepool-fqLfdx"
  login_as_non_root = false
  runtime_name      = "containerd"
  cluster_id        = "ca8ed83XXXXXXXXXXXXXXXXa55ec197"
  password          = "XXXXXXX"

  data_disks {
    size              = 120
    encrypted         = "false"
    performance_level = "PL0"
    category          = "cloud_essd"
  }

  instance_charge_type = "PostPaid"               // The billing method of the nodes. Valid values: PrePaid: subscription. PostPaid: pay-as-you-go. Default value: PostPaid.
  desired_size         = 1
  unschedulable        = false

  scaling_config {
    enable = false
  }

  image_type    = "WindowsServer"
  vswitch_ids   = [
    "vsw-bp1XXXXXXXXXXXXXXXX84",
    "vsw-bp1XXXXXXXXXXXXXXXX0a",
    "vsw-bp1XXXXXXXXXXXXXXXXXo"
  ]
  
  instance_types = [                             // The instance types. The instance types of nodes in the node pool. When the system adds a node to the node pool, the system selects the most appropriate one from the specified instance types for the node.
    "ecs.g6.large" 
  ]
  soc_enabled                   = false          // Specifies whether to enable MLPS security hardening. Valid values: true: enables MLPS security hardening. false: disables MLPS security hardening.
  internet_max_bandwidth_out    = 1              // The maximum outbound bandwidth of the public IP address of the node. Unit: Mbit/s. Valid values: 1 to 100.
  system_disk_encrypted         = false          // Specifies whether to encrypt the system disk. Valid values: true: The system disk is encrypted. false: The system disk is not encrypted.
  system_disk_size              = 40
  install_cloud_monitor         = true
  system_disk_category          = "cloud_essd"
  system_disk_performance_level = "PL0"
  image_id                      = "win2022_21H2_x64_dtc_en-us_40G_container_alibase_20240220.vhd"
  security_hardening_os         = false
}