Distributed Cloud Container Platform for Kubernetes (ACK One) This topic describes the ACK One quick-start workflow, how to use its capabilities, and where to find related documentation to help you get started with ACK One.ACK One features and references below.
Prerequisites
Distributed Cloud Container Platform for Kubernetes (ACK One) is activated with the default RAM role (see ACK One service role policies) and required Alibaba Cloud services.
Quickstart
The following figure shows the ACK One quick-start workflow.
How to use ACK One
ACK One provides the following features: registered clusters, Fleet management, Kubernetes clusters for distributed Argo workflows, and backup center.
Registered clusters
Register Kubernetes clusters deployed in data centers or on third-party platforms to ACK One for centralized hybrid cloud management.
|
Feature |
Description |
References |
|
Create a registered cluster and connect a data center or third-party Kubernetes cluster |
Create a registered cluster and connect a Kubernetes cluster deployed in a data center or on a third-party platform for centralized management in ACK One. |
|
|
Elasticity |
Scale self-managed Kubernetes clusters deployed in data centers by adding cloud compute resources. For example, manually or automatically adjust the number of Elastic Compute Service (ECS) instances or ECS bare metal instances. |
|
|
Observability |
Supports Event Center, Ingress dashboards, log collection, ARMS-based application monitoring, ARMS-Prometheus, Node Problem Detector (NPD), and Metrics Adapter. |
|
|
Security management |
Provides RAM-based authentication, RBAC authorization, cluster auditing with Simple Log Service (SLS), and cluster inspection. |
|
|
Coordinated scheduling |
Install ack-co-scheduler in a registered cluster to use ACK scheduling features for applications such as big data and AI workloads, improving resource utilization. |
Fleet management
Fleet instances are fully-managed resources in ACK for managing Kubernetes clusters across environments with a consistent cloud-native experience.
|
Feature |
Description |
References |
|
Enable Fleet management |
After enabling Fleet management, use ACK One Fleet instances to schedule and distribute workloads, applications, and configurations across clusters. |
|
|
Associated cluster management |
Associate clusters with a Fleet instance to distribute applications or workloads to them. |
|
|
GitOps |
Enable GitOps on an ACK One Fleet instance to manage application manifests and Helm charts in Git repositories with multi-cluster continuous delivery. |
|
|
Multi-cluster gateways |
ACK One multi-cluster gateways provide application disaster recovery, north-south traffic management, and zone-disaster or geo-disaster recovery for hybrid cloud and multi-cluster environments. |
|
|
MCS |
Multi-cluster Services (MCS) enables cross-cluster Service access without load balancers. |
|
|
Job distribution |
ACK One job distribution orchestrates and distributes AI workloads across multi-cluster and hybrid cloud environments. Schedule tasks to optimal clusters based on factors such as real-time resource availability and scaling demands. |
|
|
Monitoring management |
Global monitoring collects and aggregates metrics from multiple clusters and displays them on a Managed Service for Prometheus (Prometheus) dashboard. |
Kubernetes clusters for distributed Argo workflows
|
Feature |
Description |
References |
|
Create workflow clusters and obtain kubeconfig files |
Workflow clusters use a serverless architecture, run Argo workflows on elastic container instances, optimize scheduling for large-scale workflows, and use preemptible instances to reduce costs. |
|
|
Workflows |
Workflow clusters are built on open source Argo Workflows. Refer to the Argo Workflows documentation to customize workflows. |
|
|
Argo Server |
Enable Argo Server to access workflow clusters, automate workflow submission via the Argo Server API, and manage workflows with the Argo UI. |
|
|
Eventing |
Workflow clusters support eventing to automatically trigger workflows based on events. |
|
|
Observability |
Workflow clusters integrate with Prometheus for cluster metrics dashboards and SLS for workflow pod log collection and analysis. |
Backup center
The backup center backs up, restores, and migrates stateful or stateless applications, and provides disaster recovery for stateful applications across hybrid cloud and multi-cluster environments.
|
Feature |
Description |
References |
|
Enable the backup center |
The backup center backs up, restores, and migrates stateful or stateless applications, and provides disaster recovery for stateful applications across hybrid cloud and multi-cluster environments. |
|
|
Backup and restoration of applications and data |
Back up and restore stateful applications in a cluster with crash consistency, application consistency, and cross-region disaster recovery. |
|
|
Cross-cluster application migration |
Migrate stateful applications across clusters with crash consistency, application consistency, and cross-region disaster recovery. |