Prepare and test your skills

Prepare and test your skills

Evaluating use cases for Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS) in Google Cloud means matching how much management work an organization wants to keep with how much it wants Google to handle. In an on-premises environment, the organization owns every layer, including cooling, physical security, and hardware maintenance. Moving to a public cloud shifts part of that stack to Google, so the team can focus on the work that creates the most business value.
The shared responsibility model defines where the boundary falls between customer-managed layers and Google-managed layers. In a private hosting setup such as colocation, the customer outsources part of the physical infrastructure but still controls hypervisors, provisioned virtual infrastructure, and workloads running on that infrastructure. In a public cloud, the customer no longer manages the whole resource stack, and processes for provisioning, configuring, and consuming resources change. Choosing IaaS, PaaS, or SaaS sets how much responsibility stays with the customer and how much moves to Google.
Migration approaches differ in how much change they require and how much cloud optimization they deliver. The six common approaches are peer options that align with different service models:
Repurchase is the clearest SaaS use case because the organization stops managing the workload and uses a fully managed application. Rehost and replatform keep more control with the customer, while refactor, re-architect, and rebuild move toward managed platform services. Refactor takes longer than rehost and requires new skills, and re-architect and rebuild are more complex still. Rebuild can take longer than rehost or refactor and is not suitable for ready-made apps because it requires rewriting the app.
The Google Cloud Adoption Framework helps an organization assess readiness for different service models across four themes: Learn, Lead, Scale, and Secure. Each theme is placed in one of three phases that describe planning maturity:
Organizations in the Tactical phase favor SaaS or rehost moves that need less change and cause less disruption. Organizations in the Strategic and Transformational phases plan for scaling and use operational data to improve IT, which aligns with PaaS and cloud-optimized services that support modernization, scalability, and faster development.
Google Cloud services that support each model include identity services and API platforms that reduce customer management work. Cloud Identity is an identity as a service (IDaaS) platform that centrally manages users and groups and enables single sign-on (SSO) with thousands of third-party SaaS applications. Service Infrastructure (opens in a new tab) is a foundational platform for creating, managing, securing, and consuming APIs and services across organizations. Service Infrastructure is used by Google APIs, Cloud APIs, Cloud Endpoints, and API Gateway, and it provides authentication, authorization, auditing, rate limiting, analytics, billing, logging, and monitoring to service producers and consumers.
The migration path to Google Cloud is an ordered workflow that informs service-model decisions. The phases must happen in order because each phase depends on the previous one:
A four-step process flow of the Google Cloud migration path: Assess discovers the existing environment, Plan creates basic cloud infrastructure, Deploy moves workloads, and Optimize applies cloud-optimized technologies. Each step depends on the previous one.
Exam tip: Repurchase migration moves to a SaaS equivalent such as Google Workspace, while rehost and replatform keep more management responsibility with your team.
The core characteristics of Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS) differ by how much of the technology stack the customer controls and how much the provider abstracts away. Google Cloud (opens in a new tab) describes this as a spectrum where IaaS gives the customer the most control and SaaS gives the customer the least operational work. More control means more responsibility for setup and upkeep, while more abstraction means faster deployment with less maintenance.
The control-versus-abstraction spectrum is the line that separates customer-managed layers from provider-managed layers. In Google Cloud, compute means the virtualized processing power and memory needed to run applications, ranging from raw virtual machine (VM) instances to fully managed platforms. As an organization moves from IaaS toward SaaS, it controls less and the provider handles more of the underlying technology. This tradeoff guides service choice because infrastructure teams may need control, while product teams may need speed.
Infrastructure as a Service (IaaS) is the cloud service model that offers the most control over the computing environment. In IaaS, the customer manages the operating system (OS), middleware, and applications, while Google manages virtualization, servers, storage, and networking. Google Cloud IaaS solutions include Compute Engine (opens in a new tab), which creates and runs VM and bare metal instances with choices of machine series, storage services, accelerators, and guest OSs, and Google Cloud VMware Engine, which runs VMware workloads natively on Google Cloud while Google manages the underlying hardware and VMware platform. IaaS is often chosen by infrastructure architects, ML infrastructure engineers, and platform engineers who need to design custom environments from scratch.
Managed platforms and services that form Platform as a Service (PaaS) sit between IaaS and SaaS and balance control with ease of use. In PaaS, Google manages the underlying infrastructure, OS, and runtime environments so customers focus on workloads rather than maintaining servers. Google Cloud managed solutions include peer options with different levels of abstraction:
Managed platforms are often preferred by data scientists, AI researchers, and ML engineers who want to provision resources and deploy workloads quickly. Control still varies inside PaaS: GKE with Standard mode gives more infrastructure control than Cloud Run, which abstracts away all infrastructure, and GKE Autopilot mode abstracts more than Standard mode.
Software as a Service (SaaS) is the cloud service model where the provider delivers a complete, ready-to-use application over the internet. In SaaS, the customer does not manage infrastructure, OS, runtime, or application code and simply uses the application. In Google Cloud, Terraform Cloud is a SaaS application that runs Terraform in a stable remote environment, securely stores state and secrets, and connects to version control systems like GitHub, GitLab, and Bitbucket so new commits can trigger Terraform plans. Cloud Identity is an Identity as a Service (IDaaS) solution that centrally manages users and groups, letting organizations manage access and compliance without running identity infrastructure.
Exam tip: The key distinction is who manages what: in IaaS you manage the OS and up, in PaaS the provider manages the OS and runtime so you manage only your workloads, and in SaaS the provider manages everything and you just use the application.
Comparing the three models shows how management responsibility, control, and typical users change across the spectrum:
| Model | What the customer manages | What the provider manages | Example services | Typical users |
|---|---|---|---|---|
| IaaS | OS, middleware, applications | Virtualization, servers, storage, networking | Compute Engine, Google Cloud VMware Engine | Infrastructure architects, ML infrastructure engineers, platform engineers |
| PaaS | Workloads and application code | Infrastructure, OS, runtime environments | Batch, Cloud Run, Cluster Director, Cluster Toolkit, GKE | Data scientists, AI researchers, ML engineers |
| SaaS | Nothing, just uses the application | Everything, including the application itself | Terraform Cloud, Cloud Identity | End users and organizations |
The models form a continuous spectrum rather than rigid categories. The same provider can offer different abstraction levels, such as GKE Autopilot versus Standard, so service and configuration choices determine how much control the customer keeps.
The shared responsibility model (opens in a new tab) in Google Cloud describes how security, maintenance, and resource management are divided between Google and the customer. The basic rule is that the customer secures what the customer brings to the cloud, while Google protects the cloud itself. As abstraction rises from IaaS to SaaS, Google takes on more responsibility and the customer takes on less.
Responsibilities that do not change form the foundation of the shared responsibility model across all service types. The customer is always responsible for securing customer data, including how data is accessed and used. Google is always responsible for securing the underlying infrastructure that hosts cloud services. This division means the customer controls content and access decisions, while Google defends the physical and foundational systems underneath.
How responsibility shifts from Infrastructure as a Service (IaaS) to Software as a Service (SaaS) shows the effect of abstraction on daily operations. In IaaS, only hardware, storage, and network are Google's responsibility, so the customer retains control and responsibility for everything built above them. In SaaS, everything except data and its access and usage is Google's responsibility, so the customer is left mainly with data and access decisions. Platform as a Service (PaaS) sits between these endpoints, with Google managing more of the platform than in IaaS but less of the complete solution than in SaaS.
A layered stack comparison showing which layers are customer-managed versus Google-managed across IaaS, PaaS, and SaaS. In IaaS the customer manages data, applications, and OS; in PaaS only data and workloads; in SaaS only data and its access, while Google manages everything else.
| Service model | Google Cloud responsibility | Customer responsibility |
|---|---|---|
| IaaS | Hardware, storage, and network | Everything above the infrastructure, including data and what the customer brings and configures |
| SaaS | Everything except data and its access and usage | Data and its access and usage |
Moving toward SaaS transfers patching, configuration, and platform operations from the customer to Google. Moving toward IaaS keeps more configuration and operational control with the customer but leaves more maintenance work with the customer.
Exam tip: In IaaS, Google secures only hardware, storage, and network, while in SaaS, Google secures everything except customer data and its access and usage — the customer always keeps responsibility for data.
How higher abstraction reduces operational overhead explains why service choice matters for modernization and scale. Google runs its own services such as Search, YouTube, and Gmail on the same infrastructure made available to customers, so customers benefit from the same security controls and practices. Google invests in security, resources, and expertise at a scale few enterprise teams can match, which frees customers to focus on business goals and innovation. Google also offers security services that help customers secure cloud environments at scale, plus strong contractual commitments that help customers maintain control over how data is processed. Teams that choose a more managed model hand more maintenance, updates, and infrastructure protection to Google, while teams that keep a less managed model keep more flexibility but must secure the systems they build above Google's hardware, storage, and network.
Infrastructure as a Service (IaaS) gives the customer the most control because the customer manages the operating system, middleware, and applications while Google manages virtualization, servers, storage, and networking. Platform as a Service (PaaS) balances control with ease of use because Google manages the infrastructure, operating system, and runtime environments so customers focus on workloads. Software as a Service (SaaS) delivers a complete ready-to-use application over the internet where the customer does not manage infrastructure, operating system, runtime, or application code and simply uses the application.
Repurchase replaces a purchased on-premises workload with a cloud-hosted Software as a Service (SaaS) equivalent such as Google Workspace so the organization stops managing the workload. Rehost moves workloads with minimal changes and keeps more management responsibility with the customer team.
The customer is always responsible for securing customer data including how data is accessed and used while Google is always responsible for securing the underlying infrastructure. In Infrastructure as a Service (IaaS) only hardware, storage, and network are Google's responsibility so the customer retains responsibility for everything built above them, while in Software as a Service (SaaS) everything except data and its access and usage is Google's responsibility.
Last updated