A company that meets its computing needs with a hybrid cloud uses both its own on-premises data center and services from public cloud providers. The term “hybrid cloud” has come to describe a broad IT strategy that encompasses many different deployment models and business use cases, as this article will explore.
What Is Hybrid Cloud?
A hybrid cloud is a computing environment that combines hardware and software running in a company’s on-premises data center with computing capacity in a public cloud run by a third-party provider. Most organizations will establish governance, security, and privacy policies that apply across the IT systems they use, including on-premises and public cloud–based resources. The tools used to instantiate these high-level policies will vary; public cloud providers share the management burden of networks and compute and storage resources with their customers, whereas on-premises systems are managed by the organization’s internal staff.
Hybrid cloud is a widely used description, but it isn’t a term of art with a precise meaning—think of it as IT’s catchall for “some stuff in the cloud and some stuff in my data center.”
The hybrid cloud concept is a broad one that can include a variety of systems, configurations, and methods of control. More-specific terms are commonly used to describe variations that still qualify as hybrid cloud.
Distributed cloud, for example, is a model that distributes public cloud services across multiple geographic locations. This allows enterprises to take advantage of the distributed nature of public cloud infrastructure while still managing operations, governance, and updates centrally.
Multicloud is another important and growing facet of the hybrid cloud model. Companies with a multicloud strategy use two or more clouds, picking the best option for any given workload. This can mean choosing offerings from different cloud providers as well as leveraging systems in the company’s own data centers.
These and other refinements show how a hybrid cloud approach, widely accepted as a viable strategy, is being implemented using many different architectures and deployment models.
Key Takeaways
- A hybrid cloud lets a company use both its on-premises data centers and multiple clouds, connecting the environments with shared management, security, and governance policies and simplified data integration.
- A hybrid cloud can let companies tap some of the best features of the public cloud, such as accessing the most up-to-date technology and responding quickly to new business needs, while keeping the control afforded by an on-premises data center.
- Companies choose hybrid clouds instead of a 100% cloud deployment for a variety of reasons, including data residency rules, an interest in a phased cloud migration, a need for low latency, and compliance with certain regulations.
- Hybrid clouds come in a wide range of deployment models.
Hybrid Cloud Explained
A hybrid cloud connects a company’s on-premises hardware and software with computing capacity from cloud providers. Companies often engage multiple public cloud providers. These cloud and on-premises computing environments generally share some level of common management and security policies as well as the ability to integrate applications regardless of where they run.
Public cloud computing resources can include software-as-a-service applications, such as ERP or CRM apps, or infrastructure as a service and platform as a service, such as compute, storage, databases, and development environments. On-premises resources span the same gamut. For example, the on-premises data center might support legacy applications that still do their jobs but need to be integrated with a cloud application, such as a legacy billing system that needs to share data with a cloud-based CRM app. Or an on-premises ecommerce app might rely on a public cloud for bursts of extra capacity if customer demand spikes.
How Does Hybrid Cloud Work?
Cloud and on-premises computing environments work together in a hybrid cloud architecture. Hybrid cloud is a broad term, encompassing a range of cloud and on-premises software and infrastructure that are integrated in some way and exchange data.
Hybrid usually implies a shared management responsibility between the public cloud provider and in-house IT staff. For example, in a cloud setup where services are managed remotely by the provider but reside in a company’s own data center, the in-house IT team manages the power, cooling, and operations of that data center, while the public cloud provider handles the maintenance and updates of certain hardware and software depending on the services being offered—SaaS, PaaS, IaaS or some combination of the three.
Management systems can provide a high-level view of a hybrid cloud environment, letting system managers monitor the interconnected nodes. Systems are also linked with shared security, networking, and data integration policies and approaches, which the company’s IT architects must establish. System managers will still use the vendor’s tools in each environment to configure and troubleshoot each element of the hybrid cloud.
Why Use Hybrid Cloud?
Organizations prefer hybrid clouds for a variety of reasons. Some intend for the mix of cloud and on-premises resources to be temporary as they incrementally move computing workloads to one or, more likely, multiple public cloud providers. The goal is to exit their on-premises data centers entirely, using a phased approach.
Other companies plan to keep a hybrid cloud setup indefinitely. Concerns about data residency requirements and control policies are one big reason. Some companies are comfortable letting many or even most applications and databases live in the public cloud, but certain customer data or product development information, for example, might need to stay on-premises, either by regulation or company policy.
Organizations may also have applications they feel are best kept on-premises. While they may decide to move many applications to SaaS—such as ERP, human resources, and CRM apps—there are often niche or custom legacy applications that they decide to leave on-premises and that must communicate with their cloud systems.
Bandwidth concerns are another reason companies choose hybrid clouds. If they have a certain application or workload that’s particularly unforgiving of latency, and their operations aren’t located close enough geographically to a cloud data center to remove that lag, keeping the system onsite makes sense.
Benefits of Hybrid Cloud
A hybrid cloud lets a company choose what elements they like about public cloud or on-premises operations and which tradeoffs they find acceptable. The growing range of hybrid deployment model options available lets companies align this mix of cloud and on-premises services even more closely with their needs. The following are some of the benefits of a hybrid cloud:
- Increased control. Some companies aren’t comfortable ceding direct control for certain data or computing workloads to a third-party provider. Areas of concern include the physical control of storage and backup systems, decisions about infrastructure configurations, and the control of applications used to run a business process. The numerous hybrid cloud models now available mean companies can granularly address specific data and resource management needs to retain the desired control while getting many benefits of the cloud.
- Data residency. Data residency is a powerful reason for choosing a hybrid cloud. With a hybrid cloud, an organization can meet its data residency requirements by keeping data either on its premises or within a specific country’s borders while still gaining some benefits of the public cloud.
- Regulatory compliance. Beyond data residency, using public cloud environments for select operations can help companies meet certain compliance attestations, such as the Payment Card Industry Data Security Standard (PCI DSS) for processing credit cards, Health Insurance Portability and Accountability Act (HIPAA) rules for patient privacy, the General Data Protection Regulation (GDPR), and even security standards such as the US’s FedRAMP federal government standard.
- Improved security. Public cloud providers can devote far more money and specially trained people to security monitoring and updates than almost any individual company. While an organization’s IT team shares the security responsibility for data in a public cloud, working with a hyperscale provider improves the odds that a system will be patched against the latest threats. A hybrid environment lets a company tap cloud-based security advantages for a swath of workloads, even if some of their computing environment needs to stay in an on-premises data center for control, latency, or other reasons.
- Cost optimization. A hybrid cloud gives a company an always-on test case that lets them constantly compare the cost of running applications in-house versus in a public cloud. On-premises systems are capital-intensive investments, and a hybrid setup allows organizations to convert some computing costs from capital expenses to more predictable operating expenses. The public cloud element also lets a company pay only for what they use, which is especially valuable for workloads with variable demand for resources.
- Responsiveness to business needs.Public clouds let IT teams quickly add capacity to meet a new or recurring business need. For example, using a public cloud can allow an organization to add elastic capacity—so-called “cloud bursting”—to let an on-premises system handle a spike in demand, or add new development environments in minutes or hours instead of the weeks it might take in-house, or get a SaaS application up and running much faster than a conventional on-premises application could be installed and configured.
- Access to the latest technologies, including AI. Connecting public cloud services to an on-premises data center can give a team access to technology innovations. For example, public cloud services can provide the kind of high performance computing needed to train an AI model on a massive data set, without the company having to buy hardware up front that it might not need on an ongoing basis.
- Multicloud options. A multicloud environment lets a company choose the right public cloud for any given workload based on that cloud’s capabilities and cost. It gives IT leaders choice and leverage with third-party providers. A multicloud approach doesn’t require a hybrid cloud setup—a company could run all its workloads on multiple public clouds—but companies often use a multicloud environment with an on-premises data center in hybrid strategies.
- Disaster recovery and risk reduction. Using at least two environments for computing—on-premises and a public cloud—creates a level of built-in risk reduction and can make disaster recovery planning simpler.
Hybrid Cloud Challenges
An on-premises data center comes with significant fixed costs—owning or leasing a building, acquiring and maintaining hardware, and employing skilled staff to run and maintain the infrastructure and applications. Public cloud services bring their own limitations, such as concerns about data residency and control. And while a hybrid strategy offers several benefits, it also brings unique challenges, including the following:
- Data residency. Data residency rules vary, and some require that select data remain in a country or region. That adds compliance complexity as companies must track where in the public cloud regulated data is stored. The answer is often to keep sensitive information on-premises to guarantee compliance. The key is to be thoughtful about where sensitive or regulated data will reside.
- Costs: Holding on to an on-premises data center means a company bears the building, maintenance, power, cooling, and personnel costs to keep the facility running. Having an on-premises data center plus multiple cloud providers can make total technology cost assessments complicated for a given process.
- Management: Each environment will tend to have its own management tools, nomenclature, and control points. IT staff tasked with monitoring and managing systems will need to be expert in managing both the on-premises and cloud environments.
- Integration: Companies can struggle to effectively connect their on-premises environments and public clouds for interoperability and to manage the related APIs needed to keep the integration working effectively. Release management becomes more challenging as API changes in the cloud or on-premises can break integrations.
- Developer complexity: Especially for companies using DevOps tactics such as continuous integration/continuous deployment, working in a complex hybrid cloud can create complications for developers in areas such as making sure deployments go to the right cloud or on-premises environment, having sufficient capacity in the environment, and meeting testing and compliance requirements.
- Data transport: IT teams might talk about this as “data gravity,” which refers to considering where the bulk of the existing data resides. It’s often less expensive and simpler to put any applications or hardware related to data in the same location. Moving data among owned and cloud data centers takes time and may incur data egress charges.
Building a Hybrid Cloud
There are huge variations in how IT teams build their hybrid clouds. When it comes to connecting SaaS and on-premises applications, integration service providers offer a variety of connectors and integration platforms to allow for data sharing.
When it comes to connecting the networks that make up a hybrid cloud, including the on-premises data center, the IT team has two main options: a VPN, which goes over the public internet and is encrypted, or a dedicated connection, which comes with bandwidth guarantees.
Layered across these choices is the need to design with a clear understanding of what elements the in-house team controls and which are the responsibility of the cloud provider. Who’s responsible for monitoring a given service, and what’s the protocol if there’s a problem? Many areas involve a shared responsibility. An example is security, where an in-house team might set certain standards and a cloud provider might be accountable for implementing and monitoring to those standards.
Hybrid Cloud Management
Hybrid cloud management approaches vary. Major cloud providers offer hybrid management systems that let IT teams look at the performance of both their public cloud environments and their on-premises systems within certain parameters. Some in-house IT teams use a software layer not from the public cloud provider, such as VMware, to manage compute, storage, and networking across locations and clouds. Others treat Kubernetes, an open source system for managing containerized applications, as this layer.
There are also cloud management platforms created specifically to manage some aspect of setting up, running, or monitoring multiple environments. With any of these approaches, it’s critical to map workloads to specific clouds, set up service level agreements (SLAs) and security protocols for each workload, and develop a corresponding monitoring and incident response strategy.
There will be a security layer between internal and external systems, which must be factored into a management plan. Look for the ability to set up a central dashboard that offers visibility into all elements of the hybrid infrastructure—this is critical for effective management.
Hybrid Cloud Use Cases
Hybrid clouds can range from the very simple, such as an environment connecting a cloud ERP with an on-premises inventory system, to the complex—for example, a sophisticated multicloud environment running a wide range of computing workloads. Here’s a sampling of hybrid cloud use cases.
- Public cloud migration. A hybrid cloud can offer a temporary configuration as a company steadily moves all its computing architecture to a public cloud, with the long-term goal of leaving its on-premises data center. A complete move like that can take years and requires a well-thought-out hybrid strategy in the interim.
- Data residency. When data needs to stay in a particular country or region, companies are sometimes most comfortable keeping it on-premises. To do that, one approach is to use public cloud resources inside a company’s own data center.
- Application development. Companies often do app dev work and testing using on-premises environments, then move applications to a public cloud to run in production. The opposite is also common. For the transition from on-premises to the cloud to work, those environments need to be configured and behave the same for everything from APIs to operating systems. Packaging applications in containers typically addresses most environment dependencies.
- Legacy applications. While many legacy applications are being replaced by SaaS, other long-serving apps remain on-premises. A hybrid cloud lets those on-premises and cloud apps share data as needed.
- Emerging technology. A public cloud lets a company quickly benefit from the latest technology advances, such as using cloud resources for the computing power to train an AI algorithm. A hybrid environment provides access to those resources without a cost-prohibitive capital investment and lets those resources draw on training data from on-premises systems if needed.
- Application integration. Companies often need to integrate some on-premises applications—say, an inventory management or a point-of-sale system—with cloud-based ERP and financial systems.
- Unified platform. Some companies are very motivated to use a single platform, such as VMware, to run all IT infrastructure elements, and a hybrid cloud architecture lets them do this on-premises and in the cloud. Container systems such as Kubernetes can also be used to homogenize operations.
- Disaster recovery. With a hybrid cloud, on-premises and public cloud resources can be fine-tuned to meet an organization’s unique disaster recovery and business continuity needs.
- Shared service. For services such as user identity, authentication, and access governance, companies might use a cloud service that applies to both cloud and on-premises applications alike.
Hybrid Cloud Examples
- A global bank might migrate databases from conventional on-premises data centers to servers that are managed by a cloud provider but reside inside the bank’s own data centers. This setup keeps certain data on-premises so the firm can continue complying with European data protection rules. The bank could then run some applications in its conventional on-premises data center, run some in the provider-managed environment in its data center, and shift some workloads to a second public cloud.
- In contrast, a technology company that’s growing fast could run some applications on a public cloud and integrate those with on-premises instances of ERP, inventory management, and lifecycle management solutions while it takes steps into cloud native application development.
- Or consider a cybersecurity company with thousands of customers globally that relies on on-premises colocation centers for computing infrastructure and plans to continue expanding into new market segments. It might choose to run select workloads on a public cloud to gain flexibility and the ability to handle bursts of demand.
Modernize Your Business with OCI
Hybrid cloud can describe a wide range of deployment models. Which is best depends on a company’s needs. Oracle Cloud Infrastructure (OCI) is unique in its ability to meet a broad spectrum of requirements, where hybrid is just one deployment modality. OCI offers a distributed cloud approach that gives you the flexibility to deploy your workloads wherever you like—whether on-premises, in a public cloud in more than 20 countries, in a multicloud environment alongside clouds from other providers, or in an Oracle-managed cloud that sits inside your own data center. And Oracle Enterprise Manager, Oracle’s on-premises management platform, manages all your Oracle deployments, in your data center or in the cloud.
To meet the needs of companies with data residency concerns, Oracle Cloud@Customer makes it possible to access any Oracle Cloud service while keeping data physically inside your data center, allowing you to tap the advantages of the cloud and address regulatory concerns. You could even choose an OCI Dedicated Region, which puts a complete OCI cloud region in your data center, applying the degree of isolation you need to help you meet governance, regulatory compliance, and data privacy requirements.
Oracle’s government cloud services meet the needs of defense, intelligence, and civilian agencies, with environments for the United States and United Kingdom. And Oracle Alloy provides a complete cloud infrastructure platform that lets partners such as independent software vendors, systems integrators, telcos, and others become cloud providers and customize Oracle Cloud to address specific market needs.
CHROs face a chasm between employee expectations and business goals. Here’s how to bridge the gap.
Hybrid Cloud FAQs
How are hybrid cloud and multicloud different?
A hybrid cloud is a computing architecture that includes hardware and software running in an on-premises data center and in a public cloud run by a third-party provider. Multicloud describes an approach that uses two or more public cloud providers. Companies often use a multicloud strategy as part of their hybrid cloud approach, combining an on-premises data center with more than one public cloud.
Can hybrid cloud help meet data residency requirements?
A hybrid cloud strategy can help companies meet data residency requirements while getting many of the benefits of a public cloud. For example, companies might put some data in a public cloud while keeping sensitive data that falls under data residency rules in their on-premises data centers. That on-premises computing capacity might be managed by the company or by a third-party provider.
Can I still use VMware for management if I’m moving to hybrid cloud?
It depends how your IT team sets up your architecture, of course, but it’s possible to still use VMware as an IT resource management tool while moving to a hybrid cloud model using both on-premises servers and public cloud servers. Think of this approach as making the hybrid environment look and function most like the existing on-premises management environment.