Cloud Computing 101: Where Do Your Files Actually Live in the 'Cloud'?

What Is the "Cloud"?

In one sentence: the cloud is simply someone else's server. When you store a file on a cloud drive, that file does not float in the sky; it resides on a physical server inside a data center, which could be located in Beijing, Shanghai, Guizhou, or overseas. The term "cloud" originated from network diagrams where engineers used a cloud shape to represent "the network over there"—the details were unimportant, only that it ran "over there."

Why Is It Called "Cloud Computing"?

Historical Background

The concept of cloud computing dates back to the 1960s, when computer scientists envisioned computing resources being consumed on demand like water or electricity. In 2006, Amazon launched AWS (Amazon Web Services), marking the beginning of the modern cloud computing era.

Core Idea

The core idea of cloud computing is resource pooling and on-demand allocation :

Integrate the compute, storage, and network resources of massive numbers of servers into a single "pool."

Users draw resources from this "pool" as needed.

Pay only for what you use; pay nothing when you don't use it.

Analogy:

Traditional approach: digging your own well for water (buying and maintaining your own servers).

Cloud computing: connecting to a municipal water pipe (renting cloud servers).

Three Cloud Service Models

IaaS: Infrastructure as a Service

You are renting an "unfurnished apartment."

The cloud provider supplies: servers, storage, networking. You install and manage: operating system, database, applications.

Representative products:

Alibaba Cloud ECS

AWS EC2

Tencent Cloud CVM

Suitable for: enterprises with technical teams that need full control over the environment.

PaaS: Platform as a Service

You are renting a "furnished apartment."

The cloud provider supplies: servers + operating system + development runtime environment. You only need to: develop your application.

Representative products:

Google App Engine

Alibaba Cloud Function Compute

Heroku

Suitable for: developers who want to focus on writing code without managing servers.

SaaS: Software as a Service

You are renting a "hotel room."

The cloud provider supplies: the complete application. You simply: log in and use it.

Representative products:

Tencent Docs, Feishu Docs

Gmail, Outlook

Salesforce

DingTalk

Suitable for: all users, no technical expertise required.

Comparison of the Three Models

The following table summarizes the responsibilities and analogies for each model:

IaaS : You manage system, applications, and data; provider manages hardware and network. Analogy: unfurnished apartment.

PaaS : You manage applications and data; provider manages hardware, OS, and middleware. Analogy: furnished apartment.

SaaS : You manage only data; provider manages everything. Analogy: hotel.

Four Cloud Deployment Models

Public Cloud

Resources are centrally managed by the cloud provider and shared among all customers.

Pros: low cost, high elasticity, no maintenance required.

Cons: data resides with a third party, limited customization.

Representative providers: Alibaba Cloud, AWS, Azure, Tencent Cloud.

Private Cloud

An enterprise builds and operates its own cloud infrastructure for exclusive use.

Pros: secure and controllable, highly customizable.

Cons: expensive, requires dedicated operations staff.

Representative technologies: VMware, OpenStack.

Hybrid Cloud

Combines public and private clouds.

Core data stays on the private cloud.

Elastic workloads burst to the public cloud.

Balances security and flexibility.

Community Cloud

Shared by multiple organizations serving a specific community.

Key Cloud Technologies

Virtualization

The cornerstone of cloud computing. A single physical server is transformed into multiple virtual machines through virtualization.

Compute virtualization: one physical machine becomes many virtual machines.

Storage virtualization: multiple disks are pooled into a single large storage pool.

Network virtualization: one physical network is partitioned into multiple virtual networks.

Distributed Storage

Data is scattered across many servers:

Data is split into multiple blocks and distributed.

Each block has multiple replicas.

If one server fails, data is not lost.

Representative technologies: GFS (Google File System), HDFS.

Load Balancing

When massive user requests arrive simultaneously, a load balancer distributes them across different servers:

Prevents any single server from being overloaded.

Increases overall processing capacity.

Automatically fails over to healthy servers when one goes down.

Containerization

A lighter-weight isolation method compared to virtual machines:

Docker: the most popular container technology.

Kubernetes: the leading container orchestration platform.

Startup speed is fast (seconds).

Resource overhead is low.

Advantages of Cloud Computing

Cost Reduction

No need to purchase expensive hardware.

Pay-as-you-go: pay only for what you use.

No dedicated operations team required.

Elastic Scaling

Automatically scale out during traffic peaks.

Automatically scale in during low periods.

No need to provision hardware in advance.

High Availability

Multi-region backup.

Automatic failover.

99.99% availability guarantee.

Rapid Deployment

Spin up a server in minutes.

No procurement, installation, or configuration of hardware.

Supports automated deployment.

Global Reach

Cloud providers operate data centers worldwide.

Businesses can deploy services to different regions quickly.

Reduces latency and improves access speed.

Cloud Computing Use Cases

Website and Application Hosting

The most common use case: hosting websites and app backends on the cloud.

Data Storage and Backup

Cloud drives, enterprise data backup.

Big Data Analytics

Massive data processing with frameworks like MapReduce and Spark.

AI and Machine Learning

GPU cloud servers for training large models.

Internet of Things

Collecting and processing data from massive numbers of devices.

Gaming

Game servers, game streaming, and cloud gaming.

Major Chinese Cloud Providers

The following table lists the top five Chinese cloud providers by market share and their characteristics:

Alibaba Cloud : Rank 1 – most comprehensive product portfolio, largest ecosystem.

Huawei Cloud : Rank 2 – strong in government and enterprise markets.

Tencent Cloud : Rank 3 – strong in gaming and social domains.

Tianyi Cloud (China Telecom) : Rank 4 – backed by a telecom operator.

Baidu Cloud : Rank 5 – strong AI capabilities.

Conclusion

Cloud computing is not mysterious—it is simply using someone else's servers to do your work.

Key takeaways:

Cloud = someone else's server , accessed over the network.

Three service models : IaaS (rent hardware), PaaS (rent platform), SaaS (rent software).

Core advantages : elasticity, cost savings, freedom from maintenance.

Key technologies : virtualization, distributed systems, containerization.

Next time you say "store the file in the cloud," you'll know it actually lives on a server in a data center, kept safe and available by a professional cloud provider.