Configure Fleet Instances for Multi-Cluster Management - ACK One

Fleet management lets you schedule and distribute workloads, applications, and configurations across multiple clusters through a Fleet instance of Distributed Cloud Container Platform for Kubernetes (ACK One). This topic describes how to enable Fleet management and manage Fleet instances.

Prerequisites

Before you begin, ensure that you have:

An Alibaba Cloud account. Only Alibaba Cloud accounts can enable Fleet management.

Enable Fleet management

Log on to the ACK One console.

On the Create Fleet page, configure the basic information for the Fleet instance.

Note

For supported regions, see Regions that support ACK One.

Parameter	Description
Region	The region where the clusters you want to manage are deployed.
VPC	The virtual private cloud (VPC) for the Fleet instance.
vSwitch	The vSwitch for the Fleet instance.
Resource Group	The resource group to which the Fleet instance belongs. Each resource belongs to only one resource group.
Labels	Key-value pairs to attach to the Fleet instance. Keys are required and must be unique per resource (maximum 64 characters). Values are optional (maximum 128 characters). Keys and values cannot start with `aliyun`, `acs:`, `https://`, or `http://`, and are not case-sensitive. If you add a label with a key that already exists, the new label overwrites the existing one. Adding more than 20 labels to a single resource invalidates all labels on that resource.
SLB instance for API server	The Server Load Balancer (SLB) instance for the API server. Defaults to Standard I (slb.s2.small) and cannot be changed.
Create and associate EIP	Whether to expose the API server to the Internet. Disabled by default. If enabled, an elastic IP address (EIP) is associated with the internal-facing SLB instance of the API server, enabling access to the API server through the Internet. Once associated, the EIP cannot be disassociated, because clusters may already be connected to the Fleet instance over the Internet.
Enable collection of operation logs and auditing logs	Whether to collect operation and audit logs. Enabled by default.

Click Enable Fleet Management.
Wait a few minutes and go to the Fleet Information page. When the Fleet instance status changes to Running, Fleet management is enabled.

Manage Fleet instances

After enabling Fleet management, the system automatically creates a default Fleet instance. The following sections describe how to view, create, switch, delete, and connect to Fleet instances.

View Fleet instance information

On the Fleet Information page, use the following tabs to inspect your Fleet instance:

Tab	Information available
Overview	Architecture diagram, feature status, number of associated clusters, ratio of healthy clusters, CPU usage, memory usage, and node count.
Basic Information	Instance ID, region, API server endpoint, and network details.
Connect to Fleet	kubeconfig file content for both internal and Internet access.
Fleet Logs	Logs of the Fleet instance.

Create a Fleet instance

Important

To create additional Fleet instances beyond the default one, first request a quota increase from Quota Center.

On the Fleet Information page, click Create Fleet.

In the Create Fleet panel, configure the following parameters:

Parameter	Description
Fleet Name	The name of the Fleet instance.
Region	The region where the clusters you want to manage are deployed.
VPC	The VPC for the Fleet instance.
vSwitch	The vSwitch for the Fleet instance.
Resource Group	The resource group to which the Fleet instance belongs. Each resource belongs to only one resource group.
Labels	Key-value pairs to attach to the Fleet instance. Keys are required and must be unique per resource (maximum 64 characters). Values are optional (maximum 128 characters). Keys and values cannot start with `aliyun`, `acs:`, `https://`, or `http://`, and are not case-sensitive. If you add a label with a key that already exists, the new label overwrites the existing one. Adding more than 20 labels to a single resource invalidates all labels on that resource.
SLB instance for API server	Defaults to Standard I (slb.s2.small) and cannot be changed.
Create and associate EIP	Whether to expose the API server to external access. By default, the API server is not exposed to external access. If enabled, an elastic IP address (EIP) is associated with the internal-facing SLB instance of the API server, enabling access to the API server through the Internet. Once associated, the EIP cannot be disassociated, because clusters may already be connected to the Fleet instance over the Internet.
Enable collection of operation logs and auditing logs	Whether to collect operation and audit logs. Enabled by default.

Click Create.
Wait a few minutes. When the Fleet instance status changes to Running, the instance is ready.

Switch between Fleet instances

On the Fleet Information page, click the expand button next to the Fleet instance name, then select the target Fleet instance.

Delete a Fleet instance

Important

Delete all clusters associated with the Fleet instance before deleting it.

On the Fleet Information page, click Delete Fleet in the upper-right corner.

Connect to a Fleet instance with Cloud Shell

Note

Cloud Shell connects to Fleet instances over the Internet only. Make sure the API server of the Fleet instance is exposed to external access before proceeding.

On the Fleet Information page, click Cloud Shell Managed Fleet in the upper-right corner. Cloud Shell opens and automatically loads the kubeconfig file for the Fleet instance.

Run the following command to list the clusters associated with the Fleet instance:

kubectl amc get managedcluster

Expected output:

Name                                Alias              HubAccepted
c984b098b377f44b9a48c74fa55cf****   cluster2-beijing   true
cc36f4feb49824c9d840d9fd9cdf9****   cluster1-beijing   true

To manage clusters using the AMC command-line tool, see Use AMC.