Service Management and Resource Abstraction in Cloud‑Native Environments Using OAM and KubeVela

With the rapid development of cloud‑native technologies, two irreversible trends have emerged in enterprise production: the number of microservices is exploding and the underlying infrastructure and middleware become increasingly diverse.

These trends bring several challenges: cost management of services and their dependent resources, efficient deployment of numerous services, and clear responsibility attribution for each service and resource.

The article proposes solving these problems by abstracting services and their resources into a unified view, linking compute, storage, and operational configurations at the service level. This abstraction helps prevent orphan resources, improve resource utilization, and enable holistic lifecycle management.

To address the lack of higher‑level abstraction, the Open Application Model (OAM) is introduced. OAM standardizes application definitions on top of Kubernetes, treating an application as a collection of components (e.g., a Java container, a database instance) and their operational traits (e.g., deployment, logging, autoscaling). By using OAM, applications become portable across Kubernetes clusters, serverless platforms, and edge environments without modification.

Based on OAM, the open‑source cloud‑native orchestration engine KubeVela is highlighted as a practical implementation that enables developers to describe applications declaratively and let platform teams provide abstracted infrastructure services.

The internal practice is organized into four major product blocks:

Component Marketplace – where infrastructure service owners (RDS, Redis, etc.) define and register reusable components.

Application Center – a lifecycle management platform that binds services with their resource dependencies, supports cost attribution, and automates environment delivery.

Cloud‑Native Operations Platform – reorganizes operational controls (deployment, elasticity, monitoring) around the application level.

Site‑Building Center – builds complete business sites on top of the Application Center, providing workflow and chain management for end‑to‑end delivery.

These platforms enable separation of concerns: platform engineers focus on abstracting and shielding infrastructure differences, while business developers concentrate on business logic and compose required resources.

Real‑world deployments show significant benefits: for international business, automated resource topology generation and configuration‑as‑code reduce manual effort by over 30%; for domestic business, the platforms improve multi‑region resource provisioning efficiency and facilitate cost governance.

The article concludes with an invitation for readers to share their experiences using KubeVela, discuss challenges, and exchange best practices.