How KubeVela 1.1 Simplifies Hybrid‑Cloud Application Delivery

Motivation for a New Abstraction

Hybrid, multi‑cloud, and edge deployments require a consistent, infrastructure‑agnostic model for application delivery. Existing Kubernetes‑based delivery tools are tightly coupled to the underlying runtime, forcing users to manage low‑level details and rely on personal expertise, which reduces productivity and increases error risk.

Core Innovations in KubeVela 1.1

Application‑Centric Open Application Model (OAM) – Declares the full delivery lifecycle (service provisioning, observability, multi‑cluster distribution, traffic routing, rolling updates) without exposing infrastructure specifics.

Programmable Workflow Engine – Workflows are defined as directed acyclic graphs (DAG) using the CUE language, providing idempotent, step‑by‑step processes that integrate naturally with GitOps.

Infrastructure‑Agnostic Control Plane – Runs as a control‑plane component in a management cluster, making delivery decisions independent of target runtimes (containers, functions, databases, VMs, etc.).

Key Capabilities in Version 1.1

Multi‑Environment & Multi‑Cluster Delivery – Applications can be patched per environment using Kustomize‑style patches and deployed to any number of clusters, either directly or via OCM, Karmada, etc.

Declarative Workflow Support – Built‑in steps such as manual approval, rollback, data passing, and Slack/DingTalk notifications enable complex delivery pipelines without a separate CI/CD system.

Service‑Mesh Integration – An Istio plugin provides out‑of‑the‑box canary and blue‑green rollout capabilities, abstracting mesh details from the user.

Application‑Centric Cloud Resource Delivery – Terraform is embedded to provision and bind hundreds of cloud services (databases, SaaS, load balancers, etc.) as first‑class components.

GitOps‑Ready Delivery Plane – Can be used standalone or alongside Argo CD, offering a unified abstraction for CD pipelines, dependency handling, and cross‑environment promotion.

Typical Use Cases

Multi‑Environment, Multi‑Cluster Delivery

Define environment‑specific patches (Kustomize‑style) to customize component counts, configuration, or resource limits per cluster. Workflows can control the order and conditions of deployments across clusters, supporting scenarios such as development‑staging‑production pipelines, per‑customer clusters, or geographic clusters.

Declarative Delivery Workflows

A workflow consists of a series of Step objects, each representing a modular capability (e.g., apply, approve, notify, rollout). Steps are defined in CUE and can be extended with custom plugins, allowing seamless integration with existing CI systems.

Service‑Mesh‑Driven Progressive Delivery

Enable the Istio plugin, then use the built‑in Rollout step to perform canary or blue‑green releases without requiring deep Istio knowledge. The plugin abstracts Istio resources and exposes a simple declarative API.

Cloud Resource Management via Terraform

Cloud services are modeled as components or traits. Examples include:

Database or SaaS service components (e.g., Alibaba‑RDS).

Observability traits such as logging or monitoring.

Infrastructure traits such as managed Kubernetes clusters, load balancers, or NAS storage.

Terraform scripts are executed as part of the application model, enabling declarative provisioning and binding of cloud resources.

GitOps‑Style Continuous Delivery

KubeVela’s control plane can be driven by a Git repository (GitOps). It supports:

Declarative CD pipelines defined in the application model.

Automatic handling of component dependencies and topology.

Unified abstraction over existing GitOps tools (e.g., Argo CD).

Out‑of‑the‑box delivery strategies such as canary, blue‑green, and multi‑cluster promotion.

Kustomize‑style patches for environment‑specific differences without requiring users to learn Kustomize.

Getting Started

Source code and documentation are hosted at:

https://github.com/oam-dev/kubevela

Official website and docs:

https://kubevela.io

Example repositories for multi‑cluster delivery, canary rollout, and GitOps best practices are linked in the documentation.