Understanding Hybrid Cloud: Definitions, Types, Architectural Characteristics, and Its Role in New Infrastructure

Hybrid cloud lacks a precise definition, but according to NIST it can be described as an IT architecture that interconnects multiple clouds.

Based on this definition, cloud mixing can take several forms:

Public‑cloud‑to‑public‑cloud mixing

Private‑cloud‑to‑private‑cloud mixing

Public‑cloud‑to‑private‑cloud mixing

Public‑cloud‑to‑traditional‑IT mixing

Conceptually, hybrid cloud resembles multi‑cloud, yet the two differ: hybrid cloud requires dedicated links (e.g., leased lines or VPN) to connect clouds, while multi‑cloud relies on a Cloud Management Platform (CMP) for unified management; hybrid cloud focuses on inter‑cloud connectivity for application communication, whereas multi‑cloud emphasizes resource management via open APIs.

Hybrid Cloud Forms

1. Public‑cloud‑to‑public‑cloud mixing – Early development; users avoid vendor lock‑in by using tools like HashiCorp Terraform to abstract infrastructure differences.

2. Private‑cloud‑to‑private‑cloud mixing – Typically connects instances of the same private cloud within an organization to reduce operational costs; heterogeneous mixes (e.g., VMware ↔ OpenStack) are less common.

3. Public‑cloud‑to‑private‑cloud mixing – The current hot spot; combines strengths of both clouds to achieve higher efficiency.

4. Public‑cloud‑to‑traditional‑IT mixing – Involves connecting on‑premise data centers with public clouds; although some consider it outside the strict hybrid‑cloud definition, many still classify it as such.

Hybrid Cloud Architectural Characteristics

1. Elasticity – Enables rapid horizontal scaling across clouds during peak loads and automatic resource release during low demand, surpassing the elasticity of single‑cloud deployments.

2. Scalability – Unified APIs hide underlying infrastructure differences, allowing seamless integration of third‑party services and rapid expansion across heterogeneous environments.

3. Security – Hybrid environments increase security complexity; a unified security management center is needed to enforce consistent policies across disparate clouds, and public‑cloud security services (DDoS protection, WAF, etc.) can extend the security perimeter.

Hybrid Cloud Design Goals – The "Five‑Pass" Model

Design must consider five connectivity dimensions: data‑center (facility) connectivity, network connectivity, data connectivity, application connectivity, and management connectivity.

1. Facility Pass – Cloud‑managed platforms orchestrate physical resources; load can be shifted between racks or data centers via tools like GSLB.

2. Network Pass – Achieved through dedicated lines, VPNs, or SD‑WAN, enabling seamless inter‑cloud traffic.

3. Data Pass – Utilizes enterprise‑grade data sync tools for full‑ and incremental replication between public and private clouds, and unified APIs for cross‑cloud resource orchestration.

4. Application Pass – Cloud‑native and PaaS platforms (e.g., Alibaba ACK) provide standard Kubernetes capabilities, virtual nodes, and federation to run workloads across clusters.

5. Management Pass – Open APIs enable unified resource management, orchestration, monitoring, alerting, and billing across all cloud assets.

The article then links hybrid cloud to the broader "new infrastructure" concept, which comprises information infrastructure (5G, IoT, industrial internet, satellite internet, cloud computing, blockchain), integrated infrastructure (digital transformation of traditional assets), and innovation infrastructure (research platforms). It argues that cloud computing is the foundational layer of new infrastructure, driving the digital economy through the DIKW (Data‑Information‑Knowledge‑Wisdom) model.

Finally, the piece concludes that hybrid cloud is the prevailing architecture for the new‑infrastructure era, offering the scalability, cost‑effectiveness, and security needed to meet the growing computational demands of modern society.