This document defines and discusses business objectives, drivers, andrequirements, and how these factors can influence your design decisions whenconstructing hybrid and multicloud architectures.

Objectives

An organization can adopt a hybrid or multicloud architecture either as apermanent solution to meet specific business objectives, or as a temporary stateto facilitate certain requirements, such as a migration to the cloud.

Answering the following questions about your business is a good way to define your business requirements, and to establish specific expectations about how to achieve some or all of your business objectives. These questions focus on what's needed for your business, not how to achieve it technically.

• Which business goals are driving the decision to adopt a hybrid ormulticloud architecture?
• What business and technical objectives is a hybrid or multicloudarchitecture going to help achieve?
• What business drivers influenced these objectives?
• What are the specific business requirements?

In the context of hybrid and multicloud architectures, one business goal for anenterprise customer might be to expand online sales operations or markets from asingle region to become one of the global leaders in their market segment. Oneof the business objectives might be to start accepting purchase orders fromusers across the globe (or from specific regions) within six months.

To support the previously mentioned business requirements and objectives, onepotential primary technical objective is to expand the IT infrastructure andapplications architecture of a company from an on-premises-only model to ahybrid architecture, using the global capabilities and services of publicclouds. This objective should be specific and measurable, clearly defining theexpansion scope in terms of target regions and timelines.

• Heeding laws and regulations about data sovereignty.
• Reducing capital expenditure (CAPEX) or general IT spending with the supportof cloud financial management and cost optimization disciplines likeFinOps.
  • Cloud adoption can be driven by scenarios that help reduceCAPEX, like building a Disaster Recovery solution in a hybrid ormulticloud architecture.
• Improving the user experience.
• Increasing flexibility and agility to respond to changing market demands.
• Improving transparency about costs and resource consumption.

Consider your list of business drivers for adopting a hybrid or multicloudarchitecture together. Don't consider them in isolation. Your final decisionshould depend on the balance of your business priorities.

After your organization realizes the benefits of the cloud, it might decide tofully migrate if there are no constraints—like costs or specific compliancerequirements that require highly secure data to be hosted on-premises—thatprevent it from doing so.

Although adopting a single cloud provider can offer several benefits, such asreduced complexity, built-in integrations among services, and cost optimizationoptions likecommitted use discounts,there are still some scenarios where a multicloud architecture can be beneficialfor a business. The following are the common business drivers for adopting amulticloud architecture, along with the associated considerations for eachdriver:

• Heeding laws and regulations about data sovereignty: The most commonscenario is when an organization is expanding its business to a new regionor country and has to comply with new data-hosting regulations.
  • If the existing used cloud service provider (CSP) has no local cloudregion in that country, then for compliance purposes the common solutionis to use another CSP that has a local cloud region in that country.
• Reducing costs: Cost reduction is often the most common business driverfor adopting a technology or architecture. However, it's important toconsider more than just the cost of services and potential pricingdiscounts when deciding whether to adopt a multicloud architecture. Accountfor the cost of building and operating a solution across multiple clouds,and any architecture constraints that might arise from existing systems.

Sometimes, the potential challenges associated with a multicloudstrategy might outweigh the benefits. A multicloud strategy might introduceadditional costs later on.

Common challenges associated with developing a multicloud strategy include the following:

• Increasing management complexity.
• Maintaining consistent security.
• Integrating software environments.
• Achieving consistent cross-cloud performance and reliability.
• Building a technical team with multicloud skills might beexpensive and might require expanding the team, unless it's managed bya third party company.
• Managing the product pricing and management tools from each CSP.
  • Without a solution that can provide unified costvisibility and dashboards, it can be difficult to efficientlymanage costs across multiple environments. In such cases, you mightuse the Looker cloud cost management solution whereapplicable. For more information, seeThe strategy for effectively optimizing cloud billing cost management.
• Using the unique capabilities from each CSP: A multicloud architectureenables organizations to use additional new technologies to improve theirown business capability offerings without being limited to the choicesoffered by a single cloud provider.
  • To avoid any unforeseen risk or complexity, assess yourpotential challenges through a feasibility and effectivenessassessment, including the common challenges mentioned previously.
• Avoiding vendor lock-in: Sometimes, enterprises want to avoid beinglocked into a single cloud provider. A multicloud approach lets them choosethe best solution for their business needs. However, the feasibility ofthis decision depends on several factors, such as the following:
  • Technical dependencies
  • Interoperability considerations between applications
  • Costs of rebuilding or refactoring applications
  • Technical skill sets
  • Consistent security and manageability
• Enhancing the reliability and availability level of business criticalapplications: In some scenarios, a multicloud architecture can provideresilience to outages. For example, if one region of a CSP goes down,traffic can be routed to another CSP in the same region. This scenarioassumes that both cloud providers support the required capabilities orservices in that region.

When data residency regulations in a specific country or region mandatethe storage of sensitive data—like personally identifiable information(PII)—within that location, a multicloud approach can provide a compliantsolution. By using two CSPs in one region to provide resilience to outages,you can facilitate compliance with regulatory restrictions while alsoaddressing availability requirements.

The following are some resilience considerations to assess beforeadopting a multicloud architecture:

• Data movement: How often might data move within your multicloud environment?
  • Might data movement incur significant data transfer charges?
• Security and manageability: Are there any potential security ormanageability complexities?
• Capability parity: Do both CSPs in the selected region offer therequired capabilities and services?
• Technical skill set: Does the technical team have the skills required tomanage a multicloud architecture?

Consider all these factors when assessing the feasibility of using amulticloud architecture to improve resilience.

When assessing the feasibility of a multicloud architecture, it's important toconsiderthe long-term benefits.For example, deploying applications on multiple clouds for disaster recovery orincreased reliability might increase costs in the short term, but could preventoutages or failures. Such failures can cause long-term financial andreputational damage. Therefore, it's important to weigh short-term costs againstthe long-term potential value of adopting multicloud. Also, the long-termpotential value can vary based on the organization size, technology scale,criticality of the technology solution, and the industry.

Organizations that plan to successfully create a hybrid or multicloudenvironment, should considerbuilding a Cloud Center of Excellence (COE).A COE team can become the conduit for transforming the way thatinternal teams within your organization serve the business during yourtransition to the cloud. A COE is one of the ways that your organization canadopy the cloud faster, drive standardization, and maintain stronger alignmentbetween your business strategy and your cloud investments.

If the objective of the hybrid or multicloud architecture is to create atemporary state, common business drivers include:

• The need to reduce CAPEX or general IT spending for short-termprojects.
• The ability to provision such infrastructure quickly tosupport a business use case. For example:
  • This architecture might be used for limited-time projects. Itcould be used to support a project that requires a high scaledistributed infrastructure within a limited duration, while still usingdata that is on-premises.
• The need for multi-year digital transformation projects that require alarge enterprise to establish and that use a hybrid architecture for sometime to help them align their infrastructure and applications modernizationwith their business priorities.
• The need to create a temporary hybrid, multicloud, or mixed architectureafter a corporate merger. Doing so enables the new organization to define astrategy for the final state of its new cloud architecture. It's common fortwo merging companies to use different cloud providers, or for one companyto use an on-premises private data center and the other to use the cloud.In either case, the first step in merger and acquisition is almost alwaysto integrate the IT systems.

Technical drivers

The preceding section discussed business drivers. To get approved, majorarchitectural decisions almost always need the support of those drivers.However, technical drivers, which can be based on either a technical gain or aconstraint, can also influence business drivers. In some scenarios, it'snecessary to translate technical drivers into business drivers and explain howthey might positively or negatively affect the business.

The following non-exhaustive list contains some common technical drivers foradopting a hybrid or multicloud architecture:

• Building out technological capabilities, such as advanced analyticsservices and AI, that might be difficult to implement in existing environments.
• Improving the quality and performance of service.
• Automating and accelerating application rollouts to achieve a fastertime to market and shorter cycle times.
• Using high-level APIs and services to speed up development.
• Accelerating the provisioning of compute and storage resources.
• Using serverless services to build elastic services and capabilitiesfaster and at scale.
• Using global infrastructure capabilities to build global ormulti-regional architectures to satisfy certain technical requirements.

The most common technical driver for both temporary hybrid and temporarymulticloud architectures is to facilitate a migration from on-premises to thecloud or to an extra cloud. In general, cloud migrations almost always naturallylead to hybrid cloud setup. Enterprises often have to systematically transitionapplications and data based on their priorities. Similarly, a short-term setupmight be intended to facilitate a proof of concept using advanced technologiesavailable in the cloud for a certain period.

Technical design decisions

The identified technical objective and its drivers are key to making abusiness-driven architecture decision and to selecting one of the architecturepatterns discussed in this guide. For example, to support a specific businessgoal, a company might set a business objective to build a research anddevelopment practice for three to six months. The main business requirement tosupport this objective might be to build the required technology environment forresearch and design with the lowest possible CAPEX.

The technical objective in this case is to have a temporary hybrid cloud setup.The driver for this technical objective is to take advantage of the on-demandpricing model of the cloud to meet the previously mentioned businessrequirement. Another driver is influenced by the specific technologyrequirements that require a cloud-based solution with high compute capacity andquick setup.

Use Google Cloud for hybrid and multicloud architectures

Using open source solutions can make it easier to adopt a hybrid and multicloudapproach, and to minimize vendor lock-in. However, you should consider thefollowing potential complexities when planning an architecture:

• Interoperability
• Manageability
• Cost
• Security

Building on a cloud platform that contributes to andsupports open source might help to simplify your path to adopting hybrid and multicloudarchitectures.Open cloud empowers you with an approach that provides maximum choice and abstractscomplexity. In addition, Google Cloud offers the flexibility to migrate,build, and optimize applications across hybrid and multicloud environments whileminimizing vendor lock-in, using best-in-breed solutions, and meeting regulatoryrequirements.

Google is also one of thelargest contributors to the open source ecosystem and works with the open source community to developwell-knownopen source technologies like Kubernetes. When rolled out as a managed service, Kubernetes can helpreduce complexities around hybrid and multicloud manageability and security.