What Is Cloud Native? Definitions, Key Technologies, and Practical Insights

Preface

Internet has transformed how we live, work, learn, and entertain. Cloud computing has evolved from physical machines to virtual machines, containers, and now cloud‑native architectures, which are now considered the future of cloud computing.

Tracing the Origin

Different people and organizations define cloud native differently. The article reviews various discussions and project experiences to share Alibaba delivery experts' understanding of cloud native, its construction, key technologies, and practical considerations.

Pivotal Definition

Pivotal, a leader in agile development, introduced Pivotal Cloud Foundry and the Spring ecosystem. In 2013 Matt Stine first coined “cloud native.” By 2015 he identified characteristics such as the 12‑factor app, microservices, API‑centric collaboration, and resilience. In 2017 he refined the definition to six traits: modularity, observability, deployability, testability, operability, and replaceability. Pivotal now emphasizes four points: DevOps, continuous delivery, microservices, and containers.

CNCF Definition

The Cloud Native Computing Foundation (CNCF), founded by Google, RedHat and others, promotes open‑source infrastructure. It popularized Kubernetes for container orchestration and now leads cloud‑native definition. Initially (2015) it defined cloud native as containerization, microservices, and container orchestration. In 2018 it added declarative APIs and service mesh, enabling fault‑tolerant, manageable, and observable loosely‑coupled systems.

Key Takeaways

The definition of cloud native continuously evolves. Different companies (Pivotal, CNCF) and even the same company at different times provide varying definitions. Future definitions will keep changing as technology and industry needs evolve.

Organizational Perspective

Pivotal targets large enterprises with an end‑to‑end PaaS solution, focusing on culture, processes, and top‑down strategy.

CNCF serves the open‑source community and emerging businesses, emphasizing technology, toolchains, and a bottom‑up approach.

Conclusion: Pivotal pioneered cloud‑native concepts, while CNCF drives best‑practice adoption.

My View of Cloud Native (Hamlet)

Just as the Internet gave rise to “Internet‑native” thinking, cloud computing requires a “cloud‑native” mindset. This abstract thinking precedes concrete products and drives technology adoption.

Fundamental Concepts

What the Cloud Can Do

Cloud computing, built on virtualization, abstracts hardware into services (IaaS, PaaS, FaaS, DaaS) and offers pay‑as‑you‑go pricing, reshaping the IT supply chain.

IaaS (Infrastructure as a Service)

Provides basic compute, storage, networking, and security capabilities.

PaaS (Platform as a Service)

Builds higher‑level services on top of IaaS, such as databases, object storage, middleware, and application services.

Serverless

Enables running code without managing servers, offering function‑as‑a‑service, application‑as‑a‑service, and container‑as‑a‑service models.

DaaS (Data as a Service)

Delivers data‑driven services (AI, big data, speech/face recognition) that become core cloud competencies.

Building Cloud‑Native Applications

Design, development, build, deployment, delivery, monitoring, and operations must be re‑engineered. Key design considerations include:

Adopt microservices to reduce monolith complexity.

Use service‑governance and monitoring to handle communication issues.

Leverage containers for deployment scalability.

Employ Kubernetes for orchestration.

Introduce Service Mesh for non‑functional concerns.

Run Service Mesh on Kubernetes for better support.

While microservices simplify individual services, overall system complexity remains high, requiring strong architectural and operational capabilities. Cloud providers can abstract this complexity via declarative YAML/JSON pipelines, enabling “cloud‑first” solutions.

Application Delivery

DevOps bridges development and operations, enabling rapid, continuous delivery with high quality. A typical pipeline includes tools such as GitHub, Travis, Artifactory, Spinnaker, FIAAS, Kubernetes, Prometheus, Datadog, Sumologic, and ELK.

Effective DevOps requires:

Developers can deploy code without ops assistance.

Mature tooling and monitoring for fault handling and rollback.

Ownership culture where developers are responsible for production experience.

Metrics such as deployment frequency, lead time, recovery time, and failure rate meeting industry standards.

Key Characteristics of Cloud‑Native Apps

Elastic scalability with lightweight containers and immutable infrastructure.

Fault tolerance via load balancing, rate limiting, circuit breaking, and automatic failover.

Observability with fine‑grained metrics, tracing, and logging.

Stable releases using automated canary, blue‑green, and rollback mechanisms.

Ease of management through automated diagnostics and reduced manual intervention.

Seamless user experience across the entire development‑to‑production lifecycle.

Flexible billing models (pay‑per‑use, reserved, spot, etc.).

Core Technologies

Containers

Originating from chroot and LXC, containers package an application with its full runtime environment, offering consistency, portability, and immutability.

Kubernetes

Google’s open‑source container orchestration system, derived from Borg, provides declarative APIs, extensibility, and has become the de‑facto standard for large‑scale container management.

Service Mesh

Decouples business logic from cross‑cutting concerns (service discovery, routing, load balancing, rate limiting) by moving them to sidecar proxies, enabling advanced traffic management, observability, and security.

Infrastructure as Code (IaC)

Describes the full lifecycle of infrastructure (creation, scaling, replacement) in code using tools like Terraform, ROS, or CloudFormation, enabling versioning, review, testing, and reproducibility.

Cloud IDE

Provides an end‑to‑end development experience in the cloud, offering code templates, distributed compilation, AI‑assisted suggestions, static analysis, and integrated CI/CD.

Cloud‑Native Adoption Path

From a delivery perspective, the roadmap includes:

Move to the cloud as a foundation.

Build a PaaS platform (e.g., Alibaba Cloud Container Service) to abstract underlying resources.

Implement DevOps on top of PaaS for continuous integration and delivery.

Adopt microservice governance, eventually using Service Mesh.

Introduce advanced microservice management such as API portals and business middle‑platforms.

Infrastructure evolution proceeds from a single data center to multi‑data‑center, then to hybrid and multi‑cloud environments.

Conclusion

Cloud native offers powerful capabilities but introduces complexity across concepts, technologies, and expectations. Continuous learning, sharing, and practical experimentation are essential for successful digital transformation.