Prepare and test your skills

Prepare and test your skills

App and infrastructure modernization is the move from fixed on-premises servers to a cloud architecture that scales automatically, uses managed services, and supports fast updates. Modernization accelerates transformation because teams spend less time maintaining servers and fixing old code and more time improving performance and releasing new features. On Google Cloud, this change is enabled by scalable infrastructure, containerization, serverless computing, and automated operations.
Legacy applications are often large monoliths built on the .NET Framework, hosted on on-premises Windows servers, and connected to on-premises SQL Server databases. Modern cloud-based applications break that pattern by using managed services, portable containers, independently deployable microservices, and serverless scaling. The choice between the two approaches affects cost, speed, and flexibility.
| Application type | Management and portability | Updates and scaling |
|---|---|---|
| Legacy monolith | Team manages its own Windows and SQL Server servers, with Windows dependencies and licensing costs and limited deployment choices | Monolithic code is difficult to upgrade, and on-premises resources offer little agility to plan, budget, and scale |
| Modern cloud application | Managed services handle infrastructure overhead, and containers provide mobility without Windows dependencies or licensing costs | Independently deployable microservices enable a high-velocity upgrade path, and serverless architecture provides full agility to scale and budget |
Modern cloud architecture is therefore more cost-effective, efficient, and resilient than the conventional on-premises approach. The organization gains flexibility to choose where and when to deploy, including multi-regional deployments with global load balancing, autoscaling, caching with Cloud CDN, and DDoS protection through Google Cloud Armor.
The modernization path is an ordered series of steps where each step removes a limitation and prepares the workload for the next phase. Before modernizing, the organization assesses applications and their dependencies to decide which workloads can move and which must stay on-premises for legacy or regulatory reasons, while teams learn cloud capabilities. Migration then follows a fixed order from least change to most change:
A left-to-right process flow showing legacy on-premises monolith moving through Rehost to Compute Engine VMs, Replatform to Cloud SQL and containers, then Re-architect into microservices on managed platforms.
Rehost transfers server management to cloud infrastructure quickly with minimal code change. Replatform reduces operational overhead further, while re-architect removes technical debt from the legacy version but takes more time, effort, and testing.
Containerization is the practice of packaging an application and its dependencies so it can run consistently across environments without depending on a specific Windows server. Knative provides a higher-level, smaller API surface than Kubernetes, which helps teams avoid Kubernetes complexity when deploying containers. Cloud Run (opens in a new tab) follows the Knative API but runs on Google infrastructure, so no Kubernetes cluster needs to be managed.
Containers in Cloud Run are autoscaled by default and billed for the duration of each request, with deployment time in seconds. Traffic flows from users or event sources to the Cloud Run service, which scales containers up or down and routes traffic between revisions using traffic splitting. Cloud Run integrates with event and backend services, so events flow from Pub/Sub, Cloud Scheduler, or Cloud Tasks to Cloud Run, and Cloud Run connects to backends such as Cloud SQL. Choose Cloud Run for autoscaled web frontends or internal microservices triggered by events, and choose the Kubernetes-based option when the workload needs operational flexibility such as GPUs or very large container counts.
Exam tip: Rehost moves an application as-is to Compute Engine VMs, while re-architect breaks a monolith into microservices on containers and managed services.
Managed services and automated operations modernize the full application lifecycle of building, testing, deploying, and monitoring. Cloud Build (opens in a new tab) automates the build-test-deploy cycle and integrates with vulnerability scanning and Binary Authorization, so images built from unknown or insecure sources are prevented from running in the deployment environment. Cloud Deploy manages the delivery pipeline, and automated testing validates changes before deployment with unit testing, integration testing, and end-to-end testing to reduce errors and regressions.
Observability on Google Cloud is provided by Google Cloud Observability (opens in a new tab), where logs flow from applications to Cloud Logging, performance signals flow to Cloud Monitoring, and latency data flows to Cloud Trace. The application depends on these services to detect issues, tune configurations, and adjust resource allocation. Continuous improvement then builds on this foundation to accelerate innovation with new technologies such as AI, reduce costs by removing inefficiencies, enhance agility to adapt quickly, and improve decision making with data analytics products such as BigQuery and Looker.
Data democratization is the practice of making trusted, up-to-date data accessible to everyone in an organization, not just specialists. In Google Cloud, a unified data platform breaks down silos between teams, clouds, and tools so business users, developers, and data scientists work from shared intelligence. That shared access speeds up decision-making because people can analyze, visualize, and act on the same data without waiting for exports or custom engineering.
Democratized data improves decision-making by replacing isolated data silos with a single source of truth that many teams can use at once. When data is cleansed, governed, and shared on one platform, marketing, finance, and operations all analyze the same trusted version instead of conflicting copies. This shared intelligence reduces delay and error, so non-technical stakeholders can explore data directly and technical teams can focus on deeper analysis. The result is faster, more consistent decisions supported by governed access and clear data stories.
BigQuery (opens in a new tab) is Google Cloud's fully managed and serverless enterprise data warehouse for storing and analyzing large datasets. Data flows from many sources into BigQuery tables, and queries run on Google-managed compute so teams do not manage servers or capacity. BigQuery supports all data types, works across clouds, and includes built-in machine learning and business intelligence within a unified platform. Its benefits for organization-wide access include:
BigQuery also breaks down multicloud silos. BigQuery Omni (opens in a new tab) provides compute that runs locally to storage on Amazon Web Services (AWS) or Azure, so teams can query data stored in Amazon Simple Storage Service (S3) or Azure Blob Storage without moving it first. Unified governance, sharing through BigQuery sharing, and protection through Sensitive Data Protection help organizations discover, mask, and share data safely across organizational boundaries.
Cloud Data Fusion (opens in a new tab) is a fully managed, cloud-native enterprise data integration service for quickly building and managing data pipelines. Data flows from disparate sources through a pipeline where it is cleansed, prepared, blended, transferred, and transformed before loading into destinations such as BigQuery for analysis. The service provides a graphical, code-free interface with a library of preconfigured connectors and transformations, and it is powered by the open source project CDAP.
A process flow showing data flowing from sources through Cloud Data Fusion for code-free preparation, into BigQuery for storage and analysis, then to Looker and Connected Sheets for visualization, delivering trusted insights to the whole organization.
Because the platform manages infrastructure automatically, business users, developers, and data scientists can build scalable integration solutions without managing servers. BigQuery Data Transfer Service complements this by automating ingestion into BigQuery on a scheduled or ad hoc basis without writing code. Together, these integration services depend on governed storage and pipelines to deliver clean, ready-to-analyze data to everyone.
Looker (opens in a new tab) is Google's enterprise platform for data visualization and business intelligence that lets teams access, analyze, and act on a trusted version of data. Analysis flows from BigQuery datasets into Looker models and dashboards, where users perform advanced analytics, predictive analysis, and custom data applications. Looker can also connect to transactional databases in other public clouds, so insights are not limited to one environment.
Everyday tools extend that same governed data to non-specialists. Connected Sheets lets users access, analyze, visualize, and share billions of rows of BigQuery data directly from Google Sheets spreadsheets, with no additional exports and with access controls. Data Studio connects to BigQuery from the Google Cloud console for building simple graphics, reports, and dashboards that tell clear data stories. Open connectivity through ODBC and JDBC drivers, programmatic tools such as Jupyter notebooks and Vertex AI Workbench, and validated third-party tools such as Tableau and Microsoft Power BI let teams use familiar apps while querying the same warehouse. The choice depends on the job to be done:
| Tool | Primary user | Choose when |
|---|---|---|
| Cloud Data Fusion | Business users, developers, data scientists | The work is integrating and preparing data through code-free pipelines |
| BigQuery | Data analysts and data teams | The work is storing and running analytic queries on large datasets |
| Looker | Analysts and business teams | The work needs governed dashboards, advanced analytics, and custom apps |
Exam tip: Choose Cloud Data Fusion to integrate and prepare data, BigQuery to store and analyze it, and Looker to visualize and act on a trusted version of it.
People connections and trusted transactions are two ways digital operations become faster and safer in the cloud. People connections mean employees, partners, and customers can work together and reach apps and data from anywhere with consistent identity and access controls. Trusted transactions mean every interaction across systems and users is protected, verifiable, and compliant, so organizations can operate with confidence.
Google Workspace (opens in a new tab) is the collaboration suite that lets an organization bring communication and cloud workloads under one identity system. With Cloud Identity, an administrator centrally controls which corporate identities can use which Google services, such as Google Workspace for collaboration, Google Cloud for custom workloads, and analytics services. This central control establishes a base level of access, and each service then adds finer-grained permissions for specific resources and actions. The result is that people can connect to the tools they need while the organization keeps ownership of who gets access to what.
Access in Google Cloud is managed through Identity and Access Management (IAM) (opens in a new tab), which grants granular permissions to specific resources. IAM follows the principle of least privilege, where identities receive only the permissions they need for their job and nothing more. For internal apps like dashboards and content systems, Identity-Aware Proxy (IAP) helps protect access for corporate users, while Identity Platform manages customer identities for websites and customer portals with features like social sign-on and multi-factor authentication. Application identities are handled separately with Google Cloud service accounts and Kubernetes service accounts, so apps can authenticate to other apps and to the platform without leaking user credentials.
The shared responsibility model (opens in a new tab) defines the security boundary between Google and the customer. The customer is always responsible for securing what they bring to the cloud, especially their data and how it is used, while Google is responsible for protecting the cloud itself and the underlying infrastructure. This split changes with the service model, because Google manages more of the stack as the customer moves from infrastructure toward software.
| Service model | Google secures | Customer secures |
|---|---|---|
| Infrastructure as a service (IaaS) | Hardware, storage, and network | Data, applications, operating systems, and access on top of that foundation |
| Software as a service (SaaS) | Everything except data and its access and usage | Data and how data is accessed and used |
Google strengthens this model by running its own services, such as Search, YouTube, and Gmail, on the same infrastructure offered to customers. Customers therefore inherit the same physical, administrative, and technical controls that protect Google at scale. Google also states that customer data is not used for advertising or for any purpose other than delivering Google Cloud services.
Exam tip: In IaaS, Google is responsible only for hardware, storage, and network, while in SaaS, Google is responsible for everything except the customer's data and its access and usage.
Verified access is the practice of checking every access request, including requests from employees and from Google personnel. OS Login ties administrator access on virtual machines to the employee lifecycle, so when a person changes roles or leaves, access is revoked with the account and supports Google two-factor authentication against account takeover. Organization policies set guardrails across an organization, folder, or project to keep configurations secure globally.
Organization policies can enforce rules such as:
Visibility into Google's own access is provided by services that log and control administrative actions. Access Approval (opens in a new tab) requires customer approval before an authorized Google administrator can access sensitive data, and approved requests are linked to logs. Access Transparency logs administrative access by Google personnel for support and availability, while Assured Workloads helps create and monitor control packages for regulated or sovereign requirements.
Encryption in Google Cloud protects data at rest, in transit, and in use. Customer data stored at rest is encrypted by default with no action required, with options for customer-managed encryption keys (CMEKs) created in Cloud Key Management Service (Cloud KMS). Keys can be software-based or hardware-protected, rotated to meet regulatory needs, auto-provisioned with Autokey, or held outside Google infrastructure with Cloud External Key Manager (Cloud EKM). Data in transit is encrypted and authenticated when it leaves Google-controlled boundaries, including virtual machine-to-virtual machine traffic inside a Virtual Private Cloud (VPC) network, MACsec over Cloud Interconnect, IPsec over Cloud VPN, and TLS and mutual TLS for application traffic.
For data in use, Confidential Computing provides a trusted execution environment that encrypts data in memory. This family includes Confidential VM for virtual machines, Confidential Google Kubernetes Engine Nodes for containers, Confidential Dataflow for streaming analytics, Confidential Managed Service for Apache Spark for data processing, and Confidential Space for joint analysis. Compliance builds on these controls through certifications, the Compliance resource center, and Assured Workloads for industries such as finance, government, healthcare, and education. Google also maintains cyber and privacy liability insurance and strong contractual commitments around control of data and how it is processed.
Transformation on Google Cloud connects faster applications, shared data, and trusted collaboration into one operating model. Modernized infrastructure and automation let teams release improvements quickly, democratized data from BigQuery through Looker lets everyone decide from the same truth, and centralized identity with verified access and encryption lets people connect and transact with confidence.
Rehost moves Windows Server and SQL Server instances as-is to Compute Engine virtual machines with minimal code change, replatform optimizes cloud infrastructure without rewriting the application such as using Cloud SQL for SQL Server instead of self-managed SQL Server, and re-architect changes how code functions by breaking a monolith into independent microservices in containers on managed services. Rehost transfers workloads quickly, replatform reduces operational overhead further, while re-architect removes technical debt but takes more time, effort, and testing.
Choose Cloud Run for autoscaled web frontends or internal microservices triggered by events, and choose the Kubernetes-based option when the workload needs operational flexibility such as GPUs or very large container counts. Cloud Run autoscales containers by default, bills for the duration of each request, deploys in seconds, and receives events from Pub/Sub, Cloud Scheduler, or Cloud Tasks while connecting to backends such as Cloud SQL.
Use Cloud Data Fusion to integrate and prepare data through code-free pipelines, BigQuery to store and run analytic queries on large datasets, and Looker to visualize and act on a trusted version of data in governed dashboards. Cloud Data Fusion cleanses, prepares, blends, transfers, and transforms data before loading into destinations such as BigQuery for analysis, while analysis flows from BigQuery datasets into Looker models and dashboards.
In both models the customer is always responsible for securing data and how it is used, while Google secures more of the stack from Infrastructure as a service (IaaS) to Software as a service (SaaS). For Infrastructure as a service (IaaS) Google secures hardware, storage, and network and the customer secures data, applications, operating systems, and access, while for Software as a service (SaaS) Google secures everything except data and its access and usage.
A healthcare organization uses Google Workspace to foster real-time communication and collaboration across its clinical and administrative teams. Because employees frequently share sensitive patient records and financial information via files and emails, the organization needs to strengthen its digital operations with additional security and compliance controls.
Specifically, the organization must meet the following requirements:
Which Google Cloud solution should the organization implement?
Last updated