Understanding Hybrid Cloud: Types, Architecture, and Its Role in New Infrastructure

Definition of Hybrid Cloud

According to the NIST definition, a hybrid cloud is an IT architecture that interconnects multiple clouds, enabling resources from different cloud environments to communicate and work together.

Forms of Cloud Mixing

Mixing between public clouds.

Mixing between private clouds.

Mixing between public and private clouds.

Mixing between public clouds and traditional on‑premise IT.

Hybrid Cloud vs. Multi‑Cloud

Hybrid cloud requires dedicated lines or VPNs to connect the clouds, while multi‑cloud relies on a Cloud Management Platform (CMP) for management.

Hybrid cloud focuses on inter‑cloud connectivity so applications can communicate; multi‑cloud focuses on managing cloud resources via OpenAPI.

Analysis of Each Hybrid Form

1. Public‑Public Cloud Mixing

This pattern emerged early because some users do not want to depend on a single vendor. Tools such as HashiCorp Terraform abstract IaaS differences and help implement cross‑cloud deployments.

2. Private‑Private Cloud Mixing

Typically involves connecting the same private‑cloud solution within a single organization to reduce operational costs. Heterogeneous private‑cloud mixes (e.g., VMware with OpenStack) are not mainstream and are omitted from detailed discussion.

3. Public‑Private Cloud Mixing

Currently the hottest hybrid‑cloud scenario. It leverages the strengths of both public and private clouds, allowing enterprises to retain existing private‑cloud investments while benefiting from the scalability and services of public clouds.

4. Public‑Traditional IT Mixing

In this case, traditional data‑center resources are connected to public clouds via dedicated lines or VPNs. Although technically not a pure hybrid cloud, many analysts (e.g., Gartner) still classify it as such because it represents the interconnection of cloud and on‑premise infrastructure.

Key Characteristics of Hybrid Cloud Architecture

Elasticity

Hybrid cloud enhances the core elasticity of cloud computing by allowing workloads to scale horizontally across multiple clouds during peak demand and to release resources during low demand, providing greater fault tolerance and resource efficiency.

Scalability

The unified API abstracts underlying infrastructure differences, enabling rapid integration of third‑party services and allowing enterprises to expand from private clouds to dedicated clouds and public clouds (ECS, bare‑metal, VPC, etc.) without being tied to a specific virtualization platform.

Security

Because the security perimeter expands across multiple clouds, hybrid cloud demands a unified security management center to enforce consistent policies, detect incidents across heterogeneous environments, and leverage public‑cloud security services (DDoS protection, WAF, DNS) to shift the security boundary from the enterprise data‑center to the cloud.

Hybrid‑Cloud Design Goals – The “Five‑Pass” Concept

Beyond traditional reliability, safety, stability and scalability, hybrid cloud adds five connectivity goals:

Data‑Center Pass : Cloud‑managed platforms can dynamically balance workloads across racks or data‑centers, using tools like global load balancers (e.g., Alibaba Cloud’s Feitian system).

Network Pass : Physical links (metro‑net, dedicated lines, fiber) and logical gateways (VPN, SD‑WAN) ensure seamless inter‑cloud networking.

Data Pass : Enterprise‑grade data‑sync tools (e.g., Alibaba Cloud ASCM) enable full and incremental data replication between public and private clouds.

Application Pass : Cloud‑native and PaaS platforms (e.g., Alibaba Cloud ACK with virtual nodes or federation) allow applications to run across multiple clusters and clouds.

Management Pass : Open APIs provide unified resource management, orchestration, alerting, billing, and enable centralized control of VPC‑to‑VPC and VPC‑to‑on‑premise connections.

Hybrid Cloud in the Context of New Infrastructure

“New infrastructure” (新基建) consists of three layers: information infrastructure (5G, IoT, industrial internet, satellite internet, cloud computing), integrated infrastructure (deep application of AI, big data, cloud to transform traditional assets), and innovative infrastructure (research platforms, scientific facilities). Cloud computing underpins the entire stack, providing the compute power needed for data processing, knowledge extraction, and wisdom generation.

The DIKW model (Data → Information → Knowledge → Wisdom) illustrates how raw data is transformed into valuable insights, a process that increasingly relies on massive cloud‑based compute resources.

As data volumes grow, the required compute power rises, making cloud platforms the most cost‑effective solution for large‑scale processing, and hybrid cloud becomes the practical architecture for enterprises that must balance public‑cloud advantages with private‑cloud control and compliance.

In summary, hybrid cloud is the prevailing architectural pattern for the new‑infrastructure era, offering elasticity, scalability, and security while enabling enterprises to leverage both public‑cloud services and existing private‑cloud investments.