Container Service for Kubernetes:Create an ACK dedicated cluster

Move the pointer over All Resources at the top of the page and select the resource group that you want to use. After you select a resource group, virtual private clouds (VPCs) and vSwitches that belong to the resource group are displayed. When you create a cluster, only VPCs and vSwitches that belong to the specified resource group are displayed.

The name of the cluster. The name must be 1 to 63 characters in length, and can contain digits, letters, hyphens (-), and underscores (_). The name must start with a letter or digit.

These billing methods are supported: Pay-As-You-Go and Subscription. If you select the subscription billing method, you must set the following parameters:

• Duration: You can select 1, 2, 3, or 6 months. If you require a longer duration, you can select 1 year, 2 years, or 3 years.

• Auto Renewal: Specify whether to enable auto-renewal.

If you enable IPv4/IPv6 dual-stack, a dual-stack cluster is created. This feature is in public preview. To use this feature, submit an application in the Quota Center console.

Select a VPC to deploy the cluster. Standard VPCs and shared VPCs are supported.

• Shared VPC: The owner of a VPC (resource owner) can share the vSwitches in the VPC with other accounts in the same organization.

• Standard VPC: The owner of a VPC (resource owner) cannot share the vSwitches in the VPC with other accounts.

You can select Create Basic Security Group, Create Advanced Security Group, or Select Existing Security Group. For more information about security groups, see Overview.

By default, ACK automatically creates an internal-facing SLB instance that uses the pay-as-you-go billing method for the API server.

Select or clear Expose API Server with EIP. The ACK API server provides multiple HTTP-based RESTful APIs, which can be used to create, delete, modify, query, and monitor resources such as pods and Services.

• If you select this check box, an elastic IP address (EIP) is created and associated with an SLB instance. Port 6443 used by the API server is opened on master nodes. You can connect to and manage the cluster over the Internet by using a kubeconfig file.

• If you clear this check box, no EIP is created. You can use a kubeconfig file to connect to the cluster only from within the VPC and then manage the cluster.

Select a network plug-in. Flannel and Terway are supported. For more information, see Work with Terway.

• Flannel: a simple and stable Container Network Interface (CNI) plug-in that is developed by open source Kubernetes. Flannel offers a few simple features and does not support standard Kubernetes network policies.

• Terway: a network plug-in developed by Alibaba Cloud Container Service. Terway allows you to assign Alibaba Cloud Elastic Network Interfaces (ENIs) to containers. It also allows you to customize network policies of Kubernetes to control intercommunication among containers, and implement bandwidth throttling on individual containers.

  Note

  • The number of pods that can be deployed on a node depends on the number of ENIs that are attached to the node and the maximum number of secondary IP addresses that are provided by these ENIs.

  • If you select a shared VPC for a cluster, you must select Terway as the network plug-in.

  • If you select Terway, an ENI is shared among multiple pods. A secondary IP address of the ENI is assigned to each pod.

  When you set Network Plug-in to Terway, you can configure the following parameters:

  • Specify whether to enable the Assign One ENI to Each Pod feature. To use the Assign One ENI to Each Pod feature, you need to log on to the Quota Center console and submit an application.

    • If you select the check box, an ENI is assigned to each pod.

      Note

      After you select Assign One ENI to Each Pod, the maximum number of pods supported by a node is reduced. Exercise caution before you enable this feature.

    • If you clear the check box, an ENI is shared among multiple pods. A secondary IP address that is provided by the ENI is assigned to each pod.

  • Specify whether to use IPVLAN.

    • This option is available only when you clear Assign One ENI to Each Pod.

    • If you select IPVLAN, IPVLAN and extended Berkeley Packet Filter (eBPF) are used for network virtualization when an ENI is shared among multiple pods. This improves network performance. Only the Alibaba Cloud Linux operating system is supported.

    • If you clear IPVLAN, policy-based routes are used for network virtualization when an ENI is shared among multiple pods. The CentOS 7 and Alibaba Cloud Linux operating systems are supported. This is the default setting.

    • You can enable or disable IPVLAN only when you create a cluster. After the cluster is created, you can no longer enable or disable IPVLAN.

    For more information about the IPVLAN mode supported by Terway, see Work with Terway.

  • Select or clear Support for NetworkPolicy.

    • The NetworkPolicy feature is available only when you clear Assign One ENI to Each Pod. By default, Assign One ENI to Each Pod is unselected.

    • If you select Support for NetworkPolicy, you can use Kubernetes network policies to control the communication among pods.

    • If you clear Support for NetworkPolicy, you cannot use Kubernetes network policies to control the communication among pods. This prevents Kubernetes network policies from overloading the Kubernetes API server.

    For more information about how to use network policies in ACK clusters and the use scenarios of network policies, see Use network policies in ACK clusters.

  • Select or clear Support for ENI Trunking. To use the Support for ENI Trunking feature, you need to log on to the Quota Center console and submit an application. The Terway Trunk ENI feature allows you to specify a static IP address, a separate vSwitch, and a separate security group for each pod. This allows you to manage and isolate user traffic, configure network policies, and manage IP addresses in a fine-grained manner. For more information, see Configure static IP addresses, separate vSwitches, and separate security groups for a pod.

If you enable IPv4/IPv6 dual stack, you must specify an IPv6 CIDR block for Services. When you set this parameter, take note of the following items:

• You must specify a Unique Local Unicast Address (ULA) space within the address range fc00::/7. The prefix must be 112 bits to 120 bits in length.

• We recommend that you specify an IPv6 CIDR block that has the same number of IP addresses as the Service CIDR block.

For more information about how to plan CIDR blocks for a cluster, see Plan the network of an ACK cluster.

iptables and IPVS are supported.

• iptables is a mature and stable kube-proxy mode. In this mode, service discovery and load balancing for Kubernetes Services are configured by using iptables rules. The performance of this mode depends on the size of the Kubernetes cluster. This mode is suitable for Kubernetes clusters that manage a small number of Services.

• IPVS is a high-performance kube-proxy mode. In this mode, service discovery and load balancing for Kubernetes Services are configured by the IP Virtual Server (IPVS) module of Linux. This mode is suitable for clusters that manage a large number of Services. We recommend that you use this mode in scenarios where high-performance load balancing is required.

Specify the container runtime based on the Kubernetes version.

• containerd: containerd is recommended for all Kubernetes versions.

• Sandboxed-Container: supports Kubernetes 1.24 and earlier.

• Docker: supports Kubernetes 1.22 and earlier.

For more information, see Comparison among Docker, containerd, and Sandboxed-Container.

The following billing methods are supported for nodes in a node pool: pay-as-you-go, subscription, and preemptible instances.

• If you select the pay-as-you-go billing method, ECS instances in the node pool are billed on a pay-as-you-go basis. You are not charged for using the node pool.

• If you select the subscription billing method, you must set the Duration and Auto Renewal parameters.

• If you select the preemptible instances billing method, you must set the following parameter.

  Upper Price Limit of Current Instance Spec: If the real-time market price of an instance type that you select is lower than the value of this parameter, a preemptible instance of this instance type is created. After the protection period (1 hour) ends, the system checks the spot price and resource availability of the instance type every 5 minutes. If the real-time market price exceeds your bid price or if the resource inventory is insufficient, the preemptible instance is released.

  ACK supports only preemptible instances with a protection period. For more information, see Overview and Best practices for preemptible instance-based node pools.

Select the ECS instances used by the worker node pool based on instance types or attributes. You can filter ECS instances by attributes such as vCPU, memory, instance family, and architecture.

When the node pool is scaled out, ECS instances of the selected instance types are created. The scaling policy of the node pool determines which instance types are used to create new nodes during scale-out activities. Select multiple instance types to improve the success rate of node pool scale-out operations.

If the node pool fails to be scaled out because the instance types are unavailable or the instances are out of stock, you can specify more instance types for the node pool. The ACK console automatically evaluates the scalability of the node pool. You can view the scalability level when you create the node pool or after you create the node pool.

• For more information about ECS instance types, see Overview of instance families.

• For more information about node pool scalability, see Check the scalability of a node pool.

• Disable: disables security hardening for ECS instances.

• Reinforcement based on classified protection: You can enable security hardening only when you select an Alibaba Cloud Linux 2 or Alibaba Cloud Linux 3 image. Alibaba Cloud provides baselines and the baseline check feature to help you check the compliance of Alibaba Cloud Linux 2 images and Alibaba Cloud Linux 3 images with the level 3 standards of Multi-Level Protection Scheme (MLPS) 2.0. For more information, see ACK reinforcement based on classified protection.

  Important

  MLPS Security Hardening enhances the security of OS images to meet the requirements of GB/T 22239-2019 Information Security Technology - Baseline for Classified Protection of Cybersecurity without compromising the compatibility and performance of the OS images.

  After you enable MLPS Security Hardening, remote logons through SSH are prohibited for root users. You can use Virtual Network Computing (VNC) to log on to the OS from the ECS console and create regular users that are allowed to log on through SSH. For more information, see Connect to a Linux instance by using VNC.

• OS Security Hardening: You can enable Alibaba Cloud Linux Security Hardening only when the system image is an Alibaba Cloud Linux 2 or Alibaba Cloud Linux 3 image.

Valid values: Key Pair, Password, and Later.

• Configure the logon type when you create the node pool:

  • Key Pair: Alibaba Cloud SSH key pairs provide a secure and convenient method to log on to ECS instances. An SSH key pair consists of a public key and a private key. SSH key pairs support only Linux instances. For more information, see Overview.

  • Password: The password must be 8 to 30 characters in length, and can contain letters, digits, and special characters.

• Configure the logon type after you create the node pool: For more information, see Bind an SSH key pair to an instance and Reset the logon password of an instance.

ESSD AutoPL, Enterprise SSD (ESSD), ESSD Entry, Standard SSD, and Ultra Disk are supported.

The types of system disks that you can select depend on the instance types that you select. Disk types that are not displayed in the drop-down list are not supported by the instance types that you select. For more information about disks, see Overview of Block Storage. For more information about disk types supported by different instance types, see Overview of instance families.

You can select More System Disk Types and select a disk type other than the current one in the System Disk section to improve the success rate of system disk creation. The system will attempt to create a system disk based on the specified disk types in sequence.

ESSD AutoPL, Enterprise SSD (ESSD), ESSD Entry, SSD, and Ultra Disk are supported. The disk types that you can select depend on the instance types that you select. Disk types that are not displayed in the drop-down list are not supported by the instance types that you select. For more information about disks, see Overview of Block Storage. For more information about disk types supported by different instance types, see Overview of instance families.

• ESSD AutoPL disks provide the following features:

  • Performance provision: The performance provision feature allows you to configure provisioned performance settings for ESSD AutoPL disks to meet storage requirements that exceed the baseline performance without the need to extend the disks.

  • Performance burst: The performance burst feature allows ESSD AutoPL disks to burst their performance when spikes in read/write workloads occur and reduce the performance to the baseline level at the end of workload spikes.

• ESSDs provide the following features:

  Custom Performance. You can select higher PLs for ESSDs with larger storage capacities. For example, you can select PL 2 for an ESSD with a storage capacity of more than 460 GiB. You can select PL 3 for an ESSD with a storage capacity of more than 1,260 GiB. For more information, see Capacity and PLs.

• You can select Encryption for all disk types when you specify the type of data disk. By default, the default service CMK is used to encrypt the data disk. You can also use an existing CMK generated by using BYOK in KMS.

• You can also use snapshots to create data disks in scenarios where container image acceleration and fast loading of large language models (LLMs) are required. This improves the system response speed and enhances the processing capability.

• Make sure that a data disk is mounted to /var/lib/container on each node, and /var/lib/kubelet and /var/lib/containerd are mounted to the /var/lib/container. For other data disks on the node, you can perform the initialization operation and customize their mount directories. For more information, see Can I mount a data disk to a custom directory in an ACK node pool?