Why PaaS Is the Strategic Core of Cloud Computing

Cloud computing offers three service models: SaaS, PaaS, and IaaS. Among them, PaaS is the most difficult to grasp, yet it occupies a strategic core position in the cloud ecosystem.

1. PaaS is the Application Infrastructure in a Cloud Environment

Some view PaaS primarily as distributed technologies (computing, storage, databases) that virtualize many machines into a super‑computer, while others see it as a cloud service providing hosted software development tools for developers. Distributed technologies are merely enabling technologies, not the entirety of PaaS. A complete PaaS platform should offer both APaaS (e.g., GAE, Heroku) and IPaaS (Integration Platform as a Service) capabilities.

PaaS can be understood as middleware‑as‑a‑service, illustrated below:

PaaS Functions

PaaS provides the infrastructure resources needed to deploy and run applications, allowing developers to ignore underlying hardware and dynamically scale resources. A complete PaaS platform should include:

Application runtime environment

Distributed runtime environment

Various data storage options

Dynamic resource scaling

Full application lifecycle support

Development SDKs, IDEs, etc., to accelerate development, testing, and deployment

Public services via APIs (queues, storage, caching)

Monitoring, management, and precise resource metering

Integration and composite application building capabilities

Connectivity, integration, messaging, and workflow services for SOA‑style composite applications

Global functional view of PaaS:

Multi‑tenant Elasticity is PaaS’s Core Feature

PaaS characteristics include multi‑tenancy, elasticity (dynamic resource scaling), unified operations, self‑healing, fine‑grained metering, and SLA guarantees. Multi‑tenant elasticity distinguishes PaaS from traditional platforms and is its most essential trait.

Multi‑tenancy means a software system serves multiple logical, isolated tenants. Elasticity allows a system to dynamically increase or release computing resources based on demand.

Technical implementations of multi‑tenancy:

Shared‑Nothing: each tenant receives a full stack identical to on‑premise, ensuring isolation but lacking elastic resource sharing.

Shared‑Hardware: tenants share physical machines; virtual machines are the smallest unit of elastic scheduling (e.g., Microsoft Azure).

Shared‑OS: tenants share the operating system; processes become the unit of elastic scheduling, offering finer sharing at some security cost.

Shared‑Everything: a metadata‑driven model that shares all resources (e.g., force.com), achieving maximal efficiency but with high technical complexity.

2. PaaS’s Strategic Core Position

Just as traditional middleware sits at the top of the IT stack, PaaS will become the high ground of the cloud industry chain, crucial for both large private clouds and SMB‑focused application clouds.

Building enterprise private clouds with PaaS at the core

Large enterprises often have under‑utilized data centers (typically <30% CPU utilization). By pooling resources and providing dynamic, on‑demand allocation, PaaS adds business agility, higher service levels, and better resource utilization.

The PaaS‑centric private‑cloud model builds on an IaaS resource pool and adds internal cloud platforms, external SaaS operation platforms, and unified development/testing environments.

Internal cloud platform: business support platform

External SaaS platform: provide SaaS applications to external partners or customers

Development & testing environment: unified platform for developers

For a large airline group expanding into tourism, logistics, finance, and retail, PaaS‑based internal and external cloud platforms enable rapid integration of IT resources, better customer response, and innovation acceleration.

Using PaaS to Build and Operate the Next‑Generation SaaS

SMBs often lack IT expertise and capital for in‑house solutions, making SaaS attractive. However, traditional SaaS struggles with customization and lack of one‑stop service. Both global (Salesforce) and domestic SaaS vendors recognize that the future lies in PaaS‑centric architectures.

In the PaaS‑driven ecosystem, each participant gains value: SMBs receive customizable SaaS, SaaS operators achieve scale, developers can SaaS‑ify mature apps, and infrastructure providers improve utilization.

3. PaaS as a “Silver Bullet” for Future Software Development

No single technology promises order‑of‑magnitude gains in productivity, reliability, or simplicity, but PaaS is a clear direction that reshapes application delivery, deepens specialization, and decouples development from operations, dramatically boosting software delivery efficiency.

PaaS changes traditional application delivery

PaaS bridges development and operations teams

4. Conclusion

Gartner predicts a surge of new PaaS offerings in 2011, making it a dominant year for the model. PaaS is increasingly merging with IaaS, offering both development platforms and underlying infrastructure management (virtualization, OS patches, security, etc.). All stakeholders—large enterprises, SMBs, software vendors, operators, developers, and ops teams—must recognize PaaS’s strategic core and prepare for the coming era.