Cloud Digital Leader · Domain 4 of 6

Cloud Digital Leader Domain 4: Modernize Infrastructure and Applications with Google Cloud

Section 4 covers how organizations move to Google Cloud, the migration and compute vocabulary, when to use Compute Engine, GKE or serverless Cloud Run, which products run in hybrid and multicloud setups, and the business value of APIs and Apigee.

Exam weight
~18%
Official objectives
3
Estimated questions estimate
9–11
Suggested study time estimate
7–10 hrs

About 12 min read Reviewed October 2026

Overview

Section 4 is about moving and modernizing workloads. Objective 4.1 covers migration vocabulary, from workload and discovery and assessment through retire, retain, rehost, replatform, refactor and reimagine, plus compute terms such as virtual machines, containers, microservices, serverless, Spot VMs, Kubernetes, autoscaling, load balancing and managed services.

Objective 4.2 asks for the value and use cases of Compute Engine, GKE, Cloud Run and Cloud Run functions, the benefits of modern application development, and products that run across hybrid and multicloud environments: AlloyDB Omni, BigQuery Omni, GKE Enterprise, Cloud SQL and Looker. Objective 4.3 covers what an API is, how public APIs create revenue and partner opportunities, and why Apigee API Management helps.

How it's tested

Section 4 scenarios describe an application, its current state and a business goal such as speed, cost or portability. You choose a migration strategy, a compute option or an API approach. The right answer balances effort against benefit, and respects constraints such as legacy code, licensing or the need to run in several environments.

  • Migration path

    An application and a deadline are described, and you choose among retire, retain, rehost, replatform, refactor and reimagine.

  • Compute choice

    Control, portability and operations effort decide between Compute Engine, GKE, Cloud Run and Cloud Run functions.

  • APIs and portability

    A partner ecosystem, monetization goal or multi-environment need points to Apigee or hybrid and multicloud products.

Reasoning it through

A manufacturer must leave its data center in six months and runs a vendor-supplied scheduling application whose code it cannot change. Refactoring is impossible without source code, and reimagining would take too long. Rehosting the application on Compute Engine virtual machines moves it quickly with minimal change, so lift and shift is the best path for this workload.

Objectives 3

Official objectives from the Google Cloud outline, in outline order.

4.1 Describe how Google Cloud helps organizations transition to the cloudMigration terms from discovery to reimagine; compute terms from VMs to managed services.
What it means
This objective builds the vocabulary for cloud migration and compute. You explain workload, discovery and assessment, retire, retain, rehost (lift and shift), replatform (move and improve), refactor and reimagine. You also define virtual machines, containerization and containers, applications and microservices, serverless computing, Spot VMs, Kubernetes, autoscaling, load balancing and managed services.
What to understand
Discovery and assessment inventory applications and dependencies before any move. Retire turns off what is no longer needed, retain leaves an app where it is for now, rehost moves it with little change, replatform makes small optimizations such as switching to a managed database, refactor changes code to use cloud-native services, and reimagine rebuilds the app or process from scratch. A VM virtualizes hardware and runs a full operating system; a container packages an app with its dependencies and shares the host kernel, so it starts faster and moves easily. Microservices split an app into small independent services. Spot VMs use spare capacity at a large discount but can be stopped at any time. Autoscaling adjusts capacity, and load balancing spreads traffic across instances.
How it shows up
Expect a short scenario where you name the migration strategy or define a compute term that solves the stated problem.
Common confusion
Confusing replatform with refactor: replatform makes small changes, refactor rewrites code for cloud-native services.
4.2 Describe the functionality, business use cases, and business value of Google Cloud's infrastructure offeringsCompute Engine, modern app development, GKE, Cloud Run and Cloud Run functions, hybrid and multicloud products.
What it means
This objective covers Google Cloud's main compute and portability products. You explain the value of Compute Engine for running VMs, the benefits of modern application development, GKE for deploying and managing containers, serverless Cloud Run and Cloud Run functions, and products that run in multicloud and hybrid environments: AlloyDB Omni, BigQuery Omni, GKE Enterprise, Cloud SQL and Looker.
What to understand
Compute Engine offers the most control and suits legacy or specialized workloads that need a specific operating system. GKE is managed Kubernetes for running many containers with orchestration, scaling and self-healing. Cloud Run runs containers serverlessly, scaling to zero when idle so you pay only for use. Cloud Run functions run small pieces of code in response to events. Modern development brings microservices, faster releases through managed services, cost optimization, scalability and resilience, and operational efficiency. GKE Enterprise manages Kubernetes clusters consistently across Google Cloud, other clouds and on-premises. BigQuery Omni analyzes data in other clouds, and AlloyDB Omni runs AlloyDB outside Google Cloud.
How it shows up
A workload and its operational needs are given, and you pick the compute product or the hybrid and multicloud option that fits.
Common confusion
Thinking Cloud Run and GKE are the same; Cloud Run hides cluster management, GKE exposes Kubernetes control.
4.3 Describe the business value of application programming interfaces (APIs)API definition, exposing and monetizing public APIs, Apigee API Management.
What it means
This objective covers APIs as a business asset. You define an application programming interface, explain how exposing and monetizing public-facing APIs can open new business opportunities, and describe the business value of Apigee API Management.
What to understand
An API is a defined way for one piece of software to request data or actions from another without knowing its internal details. Public APIs let partners and developers build on a company's services, extend its reach and create new revenue streams, for example by charging per call. They also let legacy systems be wrapped and reused by modern apps. Apigee is Google Cloud's API management platform; it adds security, rate limiting, analytics, a developer portal and monetization features so organizations can publish and govern APIs at scale.
How it shows up
Scenarios describe sharing data with partners, monetizing services or modernizing legacy access, and ask for the API approach or Apigee.
Common confusion
Assuming an API is a product or database; it is an interface between software components.

High-yield concepts

The terms, rules and formulas this domain turns into questions most often.

  • Rehost (lift and shift)

    Move an app to the cloud with little or no change.

    Remember
    Fastest path; fits apps whose code cannot change.
    Tested as
    Deadline-driven migration scenarios.
    Trap
    Assuming it delivers full cloud-native benefits.

  • Replatform (move and improve)

    Move with small optimizations such as a managed database.

    Remember
    Some benefit without rewriting code.
    Tested as
    Moderate-effort migration questions.
    Trap
    Confusing it with refactor.

  • Refactor and reimagine

    Rewrite code for cloud-native services; rebuild the app or process entirely.

    Remember
    Highest effort and highest long-term benefit.
    Tested as
    Strategic app scenarios.
    Trap
    Choosing them under a tight deadline.

  • Retire and retain

    Shut down unneeded apps; leave some where they are for now.

    Remember
    Not every workload should move.
    Tested as
    Portfolio assessment questions.
    Trap
    Forcing every app into the cloud.

  • VM vs. container

    Full OS per instance vs. app plus dependencies sharing the host kernel.

    Remember
    Containers are lighter and more portable.
    Tested as
    Compute choice questions.
    Trap
    Thinking containers include their own OS kernel.

  • Spot VMs

    Discounted VMs using spare capacity that can be reclaimed.

    Remember
    Renamed from preemptible VMs.
    Tested as
    Batch and fault-tolerant workload cost questions.
    Trap
    Using them for critical always-on services.

  • Compute Engine

    Virtual machines on Google's infrastructure.

    Remember
    Most control; good for legacy and specialized apps.
    Tested as
    Rehost scenarios.
    Trap
    Choosing it when the team wants no server management.

  • GKE

    Managed Kubernetes for container orchestration.

    Remember
    Scales and heals containerized apps.
    Tested as
    Microservices at scale.
    Trap
    Picking it for a single simple service.

  • Cloud Run and Cloud Run functions

    Serverless containers; event-driven code snippets.

    Remember
    Scale to zero, pay per use.
    Tested as
    Spiky or event-driven workloads.
    Trap
    Looking for the old Cloud Functions name.

  • GKE Enterprise

    Consistent Kubernetes management across clouds and on-premises.

    Remember
    Replaces Anthos in the 2026 guide.
    Tested as
    Hybrid and multicloud questions.
    Trap
    Searching for Anthos in the options.

  • BigQuery Omni and AlloyDB Omni

    Analytics and database capabilities running outside Google Cloud.

    Remember
    Support data where it already lives.
    Tested as
    Multicloud data scenarios.
    Trap
    Assuming data must move into Google Cloud first.

  • Apigee API Management

    Platform to secure, manage, analyze and monetize APIs.

    Remember
    Turns APIs into a governed business channel.
    Tested as
    Partner and monetization scenarios.
    Trap
    Thinking it hosts the backend application.

Side by side

Concepts candidates mix up, compared on the point that decides the answer.

Migration strategies

ConceptMeaningKey differenceCommon trap
Rehost Lift and shift with minimal change Fastest, least benefit Expected to cut operations work sharply
Replatform Move and improve with small changes Adds managed services without rewrites Called refactor
Refactor Change code for cloud-native services More effort, more scalability Chosen for apps without source code
Reimagine Rebuild the app or process from scratch Highest effort, biggest transformation Chosen under a short deadline

Compute options

ConceptMeaningKey differenceCommon trap
Compute Engine Virtual machines Full OS control Picked by teams avoiding server work
GKE Managed Kubernetes clusters Orchestration for many containers Picked for one tiny event handler
Cloud Run Serverless containers No cluster to manage, scales to zero Assumed unable to run containers
Cloud Run functions Event-driven functions Small code units triggered by events Used for long-running stateful apps

Exam traps

Each trap is an answer that looks right. Learn the clue that should change your mind.

  1. Rewrite under deadline

    What candidates pick
    Refactor a legacy app into microservices before a data center lease ends in three months.
    Why it looks right
    Microservices sound modern and strategic.
    Why it's wrong
    Refactoring takes time and code access; rehosting meets the deadline, and modernization can follow later.

    Clue to watch for Fixed exit date and a short runway.

  2. Spot VMs for critical work

    What candidates pick
    Run the checkout service on Spot VMs to cut cost.
    Why it looks right
    Large discounts are attractive.
    Why it's wrong
    Spot VMs can be reclaimed at any time, which suits fault-tolerant batch work, not critical live services.

    Clue to watch for Always-on, customer-facing workload.

  3. Cluster for a simple service

    What candidates pick
    Deploy one small stateless web service on GKE.
    Why it looks right
    Kubernetes is the well-known container platform.
    Why it's wrong
    Cloud Run runs containers without managing a cluster and scales to zero, which is simpler for one service.

    Clue to watch for Small team, single service, variable traffic.

  4. VM for no-ops

    What candidates pick
    Choose Compute Engine for a startup that wants to avoid patching servers.
    Why it looks right
    VMs feel familiar and flexible.
    Why it's wrong
    Compute Engine leaves OS management to the customer; serverless Cloud Run removes it.

    Clue to watch for No operations staff, focus on code.

  5. Data movement for multicloud

    What candidates pick
    Copy all data from another cloud into BigQuery before analysis.
    Why it looks right
    Centralizing data seems tidy.
    Why it's wrong
    BigQuery Omni can analyze data where it resides in other clouds, avoiding movement cost and delay.

    Clue to watch for Data stays in another cloud, compliance or egress concerns.

  6. Apigee as hosting

    What candidates pick
    Choose Apigee to run a partner-facing backend application.
    Why it looks right
    Apigee is tied to APIs, so it seems to host them.
    Why it's wrong
    Apigee manages, secures and monetizes APIs in front of backends; compute products host the application itself.

    Clue to watch for Need for security, analytics or monetization of APIs.

Answer framework

Apply these steps to Section 4 scenarios about migration, compute or APIs.

  1. Check whether the workload should move at all; retire unused apps and retain those blocked by dependencies or regulations.
  2. Weigh the deadline and code access: tight timelines or vendor code point to rehost, small improvements to replatform, strategic apps to refactor or reimagine.
  3. Pick compute by control needed: Compute Engine for full OS control, GKE for many orchestrated containers, Cloud Run or Cloud Run functions for minimal operations.
  4. If workloads must run across clouds or on-premises, consider GKE Enterprise, BigQuery Omni, AlloyDB Omni, Cloud SQL or Looker.
  5. When partners or new revenue are mentioned, recommend exposing APIs and managing them with Apigee for security, analytics and monetization.

Practice question

An online ticketing startup has two developers and no operations staff. Its containerized booking service sees almost no traffic overnight but heavy spikes when concerts go on sale. Leadership wants to minimize operations work and pay only for what it uses. Which option fits best?

Want more like this? The Cloud Digital Leader practice test has questions for every domain.

Checklist

0 of 11

Tick each one you can do without notes. Your ticks stay in this browser.

Study strategy

Start with the migration ladder. Write each strategy from retire to reimagine on its own line with effort, speed and benefit, then practice sorting example applications onto the ladder. Pay attention to clues such as a fixed exit date, vendor-owned code or a strategic customer app, because they usually decide the answer.

Then learn the compute spectrum from most control to least operations: Compute Engine, GKE, Cloud Run, Cloud Run functions. For each, note who manages the operating system, how scaling works and the typical use case. Add Spot VMs as a pricing option with its reclaim risk. Update any older notes so App Engine, Anthos, Cloud Functions and preemptible VMs are replaced by the names in the August 2026 guide.

Finish with portability and APIs. Memorize the five hybrid and multicloud products the guide lists, and be ready to explain Apigee in business terms. Plan about 7 to 10 hours and mix migration and compute questions in practice sets.

FAQ

How much of the Cloud Digital Leader exam is Modernize Infrastructure and Applications with Google Cloud?

Modernize Infrastructure and Applications with Google Cloud is about 18% of the exam in the August 2026 guide, roughly 9 to 11 questions.

Is App Engine still on the exam?

App Engine is not named in the August 2026 guide. The serverless products listed are Cloud Run and Cloud Run functions.

What happened to Anthos?

The current guide no longer mentions Anthos. It lists GKE Enterprise, BigQuery Omni, AlloyDB Omni, Cloud SQL and Looker as products that support hybrid and multicloud environments.

Do I need to know Kubernetes commands?

No. You need to know what Kubernetes and containers are, why they help modernization, and when GKE is a better fit than Cloud Run or virtual machines.

Which migration term causes the most confusion?

Replatform versus refactor. Replatform moves an app with small optimizations, such as switching to a managed database, while refactor changes the code itself to use cloud-native services.

Why are APIs in a modernization section?

APIs let organizations expose existing systems to new apps and partners without rewriting them, which supports modernization and can create new revenue streams through monetization.

Other Cloud Computing certifications

Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.

Test your Domain 4 knowledge

Test whether you can quickly choose migration paths, compute products and API strategies for realistic modernization scenarios.

Take the Cloud Digital Leader practice test Check your Cloud Digital Leader readiness

Keep going