Prepare and test your skills

Prepare and test your skills

Strategic business motivations are the business reasons an organization runs parts of an application in more than one environment instead of entirely in one cloud. A hybrid architecture runs some parts of an application in Google Cloud (opens in a new tab) and other parts on-premises, while a multicloud architecture runs some parts in Google Cloud and other parts in other cloud platforms. Common motivations include reducing vendor lock-in, choosing best-of-breed services, improving resilience, meeting compliance needs, and optimizing costs.
Exam tip: Hybrid connects Google Cloud with on-premises, while multicloud connects Google Cloud with other cloud providers.
Resilience and disaster recovery is the use of a second environment as a standby so work can continue if the primary environment fails. In a hybrid design, a mission-critical on-premises application backs up its data to Google Cloud and keeps a replica there, and the replica is activated in Google Cloud if the on-premises system goes down. In a multicloud design, Google Cloud can be the primary site and another cloud can hold a passive replica, and the external replica restores production if Google Cloud goes down. Backup frequency and whether the replica is active or passive depend on the recovery time objective (RTO) and recovery point objective (RPO).
Access to best-of-breed capabilities means using the strongest service from each provider while keeping applications where they already run. Organizations often want advanced Google Cloud capabilities such as low-cost unlimited archive storage, artificial intelligence (AI) and machine learning (ML), and data warehousing and analytics with BigQuery (opens in a new tab). In a hybrid pattern, data flows from the on-premises application to Google Cloud for analysis, and in a multicloud pattern, data flows from applications on other clouds into Google Cloud for processing. This lets the organization pick the best service for each job without moving everything.
Development control and tiered topologies describe how hybrid architectures split where code is built and where it runs. Development environments stay on-premises, and a continuous integration/continuous delivery (CI/CD) pipeline pushes updates from on-premises development to Google Cloud for production. This gives the organization control over development while gaining cloud scalability, cost optimization, and reliability. A tiered topology, also called a split-stack deployment, places the application frontend in Google Cloud and the backend on-premises, and cloud bursting sends overflow traffic to Google Cloud when load on the on-premises application reaches peak capacity.
Compliance and monitoring patterns separate sensitive data and centralize visibility across environments. The separate application and operations pattern segregates sensitive application data from less sensitive operations data to meet compliance rules for sensitive data. The single pane of glass pattern centralizes monitoring and logging so data flows from on-premises and cloud sources into one cloud-hosted repository. That central repository provides one consistent view, one place to manage storage and retention, and centralized access control and auditing, but data in transit to the repository still needs security.
Cost and operational tradeoffs are the extra expenses and management work that come with running in more than one environment. Hybrid and multicloud architectures often cost more than running entirely in Google Cloud because data may be stored redundantly in each environment, which raises storage and data transfer costs. Additional costs come from underutilized redundant resources, hybrid or cross-cloud networking such as Cross-Cloud Interconnect (opens in a new tab), and monitoring resources across multiple environments. Organizations should compare cost per unit of work, such as users served or requests processed, and can use the Google Cloud Pricing Calculator to compare Google Cloud resource costs.
Operational complexity is higher because security features and tools are not standard across platforms. Security administrators need skills for each platform, and provisioning resources across platforms takes more effort. Terraform can reduce provisioning effort, and Google Kubernetes Engine (GKE) attached clusters can manage containerized applications across clouds. The organization must weigh competitive pricing against personnel costs and the overhead to design, build, and maintain the deployment.
Comparing hybrid and multicloud motivations shows how what runs where shapes the business reason for each strategy.
| Strategy | What runs where | Common business motivations | Example use cases |
|---|---|---|---|
| Hybrid | Some parts in Google Cloud, other parts on-premises | Disaster recovery, development control, enhancing on-premises apps, tiered split-stack | On-premises app with Google Cloud recovery site; on-premises development with CI/CD to cloud; BigQuery, AI, and ML for on-premises data; cloud bursting; frontend in Google Cloud and backend on-premises |
| Multicloud | Some parts in Google Cloud, other parts in other clouds | Disaster recovery, enhancing apps on other clouds, best-of-breed services | Google Cloud primary site with another cloud as recovery site; archive storage, AI, ML, and BigQuery for apps on other clouds |
The table shows that hybrid focuses on linking on-premises systems to the cloud, while multicloud focuses on combining Google Cloud with other cloud providers.
Compliance and data sovereignty requirements are legal and regulatory rules that control where data may be stored and how sensitive systems may connect to the cloud. Hybrid and multicloud architectures help meet these rules by keeping some data on-premises, placing other data in a required region, and linking environments through controlled networks. The final choice depends on business priorities, legal restrictions, and cloud benefits.
Data sovereignty and data residency are laws about where data can be hosted. A common case occurs when an organization expands into a new country and must follow new data-hosting rules for sensitive data such as personally identifiable information (PII). When regulations require that PII stay within a specific location and the current cloud service provider (CSP) has no local cloud region there, a common compliant solution is to use another CSP that does have a local region. The data therefore stays in the required country while the organization still uses cloud services.
Exam tip: Data residency rules can require sensitive data such as PII to stay in a specific country or region. If the current CSP has no local region there, using another CSP with a local region is a common compliant solution.
Hybrid cloud and on-premises sensitive data describes keeping highly secure data in a private environment while connecting it to public cloud resources. Compliance requirements can prevent full migration when highly secure data must stay hosted on-premises, and industry regulations and privacy laws can restrict moving certain applications that hold sensitive data. In that case, the organization keeps sensitive data and legacy systems on-premises and exposes only needed capabilities to cloud applications. A common pattern is to modernize legacy services with API interfaces, which unlocks them for consumption by new cloud services, and with an API management platform such as Apigee (opens in a new tab), this integration adds security, analytics, and scalability with minimal application changes.
Multicloud for data residency and resilience uses more than one CSP to stay compliant while staying available. For example, if regulations require sensitive data to stay in a specific country, using two CSPs in that same region can provide a compliant solution. If one CSP region goes down, traffic can be routed to the other CSP in the same region, but this works only if both providers support the required capabilities in that region. Before adopting this approach, the organization should assess key considerations:
Secure integration of legacy systems with public cloud connects sensitive on-premises systems to cloud resources without exposing them to uncontrolled traffic. The organization first defines application communication requirements, including the optimal network and security design. A mirrored network pattern helps design the architecture so systems in different environments do not communicate directly, and any required cross-environment communication happens in a controlled manner. A zero trust service mesh helps control and maintain that separation between environments. During deployment, the organization establishes connectivity between Google Cloud and the other environments based on the selected design and network type, and the landing zone (opens in a new tab) design defines the Google Cloud organization, projects, and resource hierarchy for the integration.
Governance and compliance assessment ensures that hybrid and multicloud designs actually meet legal and regulatory restrictions. An approval and governance process controls what the company is allowed to build in the cloud and what data can be used for testing. It can also prevent use of cloud features in development and testing that do not exist in on-premises production. During assessment, the organization identifies data privacy requirements, compliance requirements, consistency requirements, and solutions across clouds. Migration risks include the need to comply with existing legal restrictions, so these factors shape which workloads are migrated or operated in each environment.
Analyze technical use cases for resilience and performance means understanding how running the same capability in more than one environment keeps systems available and fast. Redundant deployment across on-premises data centers, Google Cloud, and other clouds increases capacity, agility, and resiliency. Three technical use cases show this value most clearly: disaster recovery planning, cloud bursting for peak demand, and reducing latency through geographic distribution.
Disaster recovery planning keeps a standby system ready in a second environment so users can be rerouted when the primary environment fails. The standby system must run the same application versions and configurations as the primary, so standby deployment is integrated into the CI/CD pipeline and artifact repositories are designed so they do not become a single point of failure. All workloads are made portable where feasible, often with containers and Kubernetes, and Google Kubernetes Engine (GKE) (opens in a new tab) simplifies building and operating those portable workloads across environments.
Traffic shifts from the primary environment to the recovery environment with DNS and load balancing. Key practices for reliable failover include:
Exam tip: Configure DNS with a short time to live value for disaster recovery, so clients pick up the new address quickly and fail over to the standby system.
Handling peak demand with cloud bursting means the base workload runs in one environment while extra demand overflows to another environment, usually the public cloud. The same application is deployed redundantly across multiple environments to add capacity without rebuilding the primary data center. When demand spikes, new capacity starts in the cloud and takes part of the load, then it shrinks again when demand falls. Bursting works only when environments are consistent and connected, so workload portability lets the burst copy run the same code as the base system and consistent CI/CD deployment keeps versions aligned. High-volume links such as Cloud Interconnect or Cross-Cloud Interconnect help when large amounts of data move between the base and burst environments, and load balancers split traffic between them based on weight or other metrics.
Reducing latency through geographic distribution means placing services physically close to users, data, or other services that need fast responses. Some services must run in a specific location because end users are sensitive to delay, because the workload needs fast access to a downstream service, or because a large data set cannot be moved and the application must run near that data. Running the same service across multiple zones, regions, or providers also provides higher availability when one location has problems. Placement decisions balance isolation, location, and scale needs, where isolation separates control planes and data planes, separates production from development and testing, and tiers critical and less-critical services. Combining multiple regional load balancers with DNS routing policies creates global load-balancing architectures that direct each client to the closest healthy environment in a hybrid or multicloud setup.
A hybrid architecture runs some parts of an application in Google Cloud and other parts on-premises, while a multicloud architecture runs some parts in Google Cloud and other parts in other cloud platforms. Common motivations for both include reducing vendor lock-in, choosing best-of-breed services, improving resilience, meeting compliance needs, and optimizing costs.
Cloud bursting runs the base workload in one environment and overflows extra demand to another environment, usually the public cloud, then shrinks the extra capacity when demand falls. It requires consistent and connected environments, so workload portability, consistent continuous integration/continuous delivery (CI/CD) deployment, high-volume links such as Cloud Interconnect or Cross-Cloud Interconnect, and load balancers to split traffic are used.
Hybrid and multicloud architectures help meet legal and regulatory rules by keeping some data on-premises, placing other data in a required region, and linking environments through controlled networks. For example, when regulations require personally identifiable information (PII) to stay within a specific location and the current cloud service provider (CSP) has no local region there, a common compliant solution is to use another CSP that does have a local region.
An enterprise hosts its core business applications inside an on-premises data center with fixed hardware capacity. During seasonal peak demand periods, user traffic surges beyond what the local servers can handle. To maintain availability without purchasing permanent hardware, the company configures its environment to dynamically route excess workloads to Google Cloud instances on demand.
Which hybrid cloud use case does this scenario describe?
Lift-and-shift migration
Active-passive disaster recovery
Cloud repatriation
Cloud bursting
Last updated