Container Service for Kubernetes:Getting started with workflow clusters - Container Service for Kubernetes

Workflow clusters are deployed on top of a serverless architecture and run Argo Workflows on elastic container instances. By optimizing cluster parameters and using preemptible instances, workflow clusters keep large-scale pipeline execution efficient, elastic, and cost-effective.

Getting started

After completing the steps in this guide, you will have a running workflow cluster and a submitted workflow.

Basic features

Task	Reference
Create a workflow cluster and get its kubeconfig file	Create a workflow cluster
Create a workflow	Create a workflow
Access the Argo Server console	Enable Argo Server for a workflow cluster and Enable Internet access for Argo Server

Advanced features

Task	Reference
Manage workflow clusters	Access the workflow cluster console through a custom domain name, Use ApsaraDB RDS for MySQL to migrate large workflows, and Modify the configuration of a workflow cluster
Manage workflows	Run workflows on a specified type of ECS instances, Use volumes, Configure artifacts, Configure a workflow to pull images from the Container Registry Enterprise Edition instance without a password, and Use a DingTalk chatbot to send event notifications
Set up eventing	Enable eventing, Use Simple Message Queue (formerly MNS) to trigger event-driven workflows, and Upload objects to OSS to trigger workflows
Enable observability	Enable Managed Service for Prometheus and Configure Simple Log Service

Billing

Workflow clusters have no cluster management fee. If you use other Alibaba Cloud services alongside workflow clusters, those services are billed according to their own pricing. For details, see ACK One billing.

What's next

To learn about the features, benefits, network design, and how workflow clusters work, see Overview of Kubernetes clusters for distributed Argo workflows.