Hybrid Cloud Concepts
What is Private Cloud?
A private cloud is a cloud-based infrastructure provisioned for the use of one organization. Unlike private clouds, a private cloud is completely isolated from other organizations. There are usually a number of units within the organization utilizing the private cloud, each referred to as a consumer.
Contrary to popular belief, private clouds aren't only deployed on premises. Private clouds might be owned by one organization and managed by another organization (a service provider) in a different physical location.
The National Institute for Standards and Technology have defined clear criteria private clouds must meet, including:
- Provide on-demand self-service of computing resources
- Make resources accessible from multiple types of devices
- Pool resources and serve them in a multi-tenant fashion
- Have the ability to automatically and elastically scale resources
- Evaluate and measure the use of resources by consumers
A computing system that does not meet these criteria is not considered a private cloud.
What is Multicloud?
A multicloud is composed of two or more cloud vendor platforms. This strategy enables organizations to leverage cloud resources offered by multiple vendors. A multicloud can be a combination of infrastructure, platform, or software as a service (IaaS, PaaS or SaaS).
For example, you can use an email service from one provider (IaaS), a customer relationship management (CRM) system with a built-in development platform from another provider (PaaS), and on-demand computing infrastructure from yet another provider's service (IaaS).
Multiple clouds enable organizations to choose the right platform for each workload, instead of trying to fit all workloads into the same environment. Different mission-critical workloads have different requirements for performance, high availability, physical data location, scalability, compliance, and so on. Cloud services from a specific vendor could meet those requirements better than others.
A multicloud environment gives an organization the freedom to use the services and infrastructure most appropriate for each workload or application, and also change the allocation of workloads to clouds as needs change or the cloud services themselves evolve.
Learn more in our in-depth guide to multicloud architecture
Hybrid Cloud vs Multi Cloud
To some extent, the distinction between hybrid cloud and multi-cloud is semantic, and in many cases the two terms can be safely interchanged. However, a common distinction is that:
- Hybrid clouds include a combination of public cloud and local or hosted private cloud.
- Multiclouds might use any type of cloud. They may or may not include an on-premise private cloud.
Apart from this basic difference, here some more aspects in which hybrid and multicloud models may differ.
There is another difference to keep in mind when comparing multicloud and hybrid cloud models. In a multi-cloud configuration, individual clouds may not be integrated with each other, and in fact they are commonly not integrated (because organizations use different clouds for different workloads). Whereas in a hybrid cloud, the local private cloud is almost always integrated, to some extent, with the public cloud.
A hybrid cloud tightly integrates private resources and public clouds, with orchestration providing a tight workflow between systems.
The private cloud portion of the hybrid architecture must be compatible with one or more public clouds that incorporate it. The hypervisor and cloud software layer must be compatible with the selected public cloud, to ensure interoperability between APIs and services.
Hybrid clouds are often used for specific tasks, such as running internal workloads, and can perform “cloud bursting”, shifting excess demand to the public cloud during periods of high computing demand (see cloud bursting below).
Multi-cloud computing, by contrast, is a comprehensive strategy that enables organizations to leverage the management and billing of multiple cloud services. The two methods are not mutually exclusive. By definition, a hybrid cloud can comprise multiple clouds, and multiple clouds can be hybrid clouds.
In terms of cost, when comparing multi-cloud and hybrid cloud environments, hybrid cloud may be more expensive because private cloud components require additional infrastructure and bandwidth, which usually require an upfront capital investment. Integrating the local data center with the cloud infrastructure also has additional cost.
Multicloud environments that mainly rely on public cloud platforms can be more cost-effective, at least in terms of upfront investment and setup. Public cloud services are typically billed per actual usage and require a much lower level of configuration and development.
However, a multicloud based on public cloud services might cost much more in the long run, because an organization needs to pay per use for every element of the computing stack, from compute, to storage, to networking and data transfer.
What is Hybrid IT?
Hybrid IT is a combination of on-premises and cloud-based services that together forms the IT environment. It can be used as a temporary solution as your business migrates to the cloud, or as a long-term solution if there is a compelling need to maintain some resources on-premises.
A hybrid IT team is mainly made up of local experts, who manage the private cloud, while configuring, managing and monitoring the public cloud elements. Among the key activities performed by hybrid IT teams are:
- Adoption planning—defining service level agreements (SLA), costs, operational procedures and migration strategies for private and public clouds.
- Data security—performing security audits and ensuring security requirements are met across public and private clouds.
- Optimization—monitoring and managing the environment to ensure resources are used optimally and workloads are allocated in the best way possible between public and private clouds.
Learn more in the in-depth guide to Hybrid IT
What is Cloud Bursting?
Cloud bursting is a configuration established to meet peak IT requirements. When consumers using the cloud reach 100% of their resource capacity, overflow traffic is routed to the public cloud, so there is no disruption to service.
When using cloud bursting, organizations must take into account security and compliance requirements, latency, load balancing, and platform compatibility between private and public clouds.
Cloud bursting has several important benefits:
- Flexibility—helps organizations serve unusual demand without having to buy and configure additional infrastructure.
- Self service—enables scaling up to serve peak demand automatically, without requiring manual intervention from IT.
- Cost savings—organizations pay only for actual excess capacity, and do not need to anticipate fluctuating demand and spend on extra capacity, which in many cases may remain unutilized.
- Optimal use of resources—cloud bursting can also be used to move non-critical workloads from local resources to public cloud resources. This helps make more space for critical business applications.
What is SD-WAN?
Software-defined WAN (SD-WAN) technology employs software-defined networking (SDN) to distribute network traffic throughout a wide area network (WAN). It uses pre-configured policies to automatically determine how to effectively route application traffic between data center sites and branch offices.
SD-WAN provides redundancy for WAN connections. It automatically fails over to a secondary path when the primary one is unavailable or fails. SD-WAN can also help improve network and application performance by applying load balancing across several connections.
SD-WAN is available on-premises or as a cloud-based SD-WAN solution. Cloud SD-WAN extracts the controller from the data center and situates it in the cloud. As a result, it introduces more scalability and network flexibility, and improved management.
SD-WAN is particularly useful for multi-cloud deployments because it enables using policies to direct traffic to the right location across private clouds, SaaS, and IaaS. To enhance multi-cloud security, SD-WAN supports microsegmentation to enable isolation of specific traffic flows.
Learn more in our guide to SD-WAN
What are the Benefits of Hybrid Cloud?
While hybrid clouds can be complex to set up, they carry significant benefits for organizations. Key benefits include cost savings, agility, and compliance.
Many organizations set up hybrid clouds to save on IT costs. However, the hybrid cloud infrastructure itself requires a lot of capital, equipment and human resources for installation and maintenance.
The key to saving costs is optimizing usage of the hybrid environment. Private clouds can provision resources like a public cloud, but unlike a public cloud, it is limited to the resources owned by the individual organization. When local demand exceeds capacity, the organization can leverage public cloud resources.
Another option is to use the public cloud for ad hoc, experimental, or general purpose workloads that do not need to run on-premises. Expensive private cloud resources can be reserved for mission-critical workloads, or workloads that will run most efficiently or cost effectively on-premises.
A hybrid cloud also makes it easier to divide IT expenses into capital and operating costs: the private cloud element is mainly a capital expense, while the public cloud element is an operating expense.
Agility is a key premise for cloud computing. Private clouds provide organizations with configurable and scalable agility, but the amount of resources available in a physical data center is still limited.
Conversely, the public cloud allows users to quickly deploy compute and storage instances and related services without being limited by resources. However, moving local workloads to the public cloud typically requires at least some pre-migration work. And public cloud services may have higher ongoing operating costs than local infrastructure.
A hybrid cloud provides both types of agility: highly configurable agility, and unlimited elastic scalability. This benefit can be enhanced by ensuring consistency. If your private cloud offers the same instance types and services as the public cloud of your choice, you can more easily create, transport, and scale your workloads and resources to the cloud resource that is most appropriate or cost effective.
A key advantage of public clouds is their global nature. Most public cloud services can support your computing, storage and networking needs from anywhere in the world. However, complexity can arise when there is a need to store or process data in specific geographical locations (data sovereignty). For some multinationals, this complicates the transition to a pure public cloud.
Hybrid clouds enable enterprises to operate sensitive workloads in a private cloud and move data back and forth between public clouds in response to changing regulatory conditions or changing requirements of specific workloads. For example, organizations can collect and organize personally identifiable customer data in a private cloud and then send it to public cloud applications, anonymized, for processing or analysis.
Hybrid Cloud Strategy
Here are some important considerations for building your organization's hybrid cloud strategy.
Choose a Consistent Architecture
To build a hybrid cloud, organizations invest in hardware, commercial software, and development work. What's unique about hybrid cloud infrastructure is that you can leverage existing investments while providing a foundation for modernization.
To achieve a good level of flexibility and portability, organizations should incorporate consistency into their architecture while adopting a public cloud. Standardizing the operating environment across private and public clouds can help reduce the complexity of cloud migration, while ensuring the business runs on existing foundations.
Decide on an Orchestration Strategy
Orchestration connects tasks across the infrastructure, creating cohesive workflows. The lack of effective orchestration between on-premises and public cloud systems can lead to loss of connectivity to key business applications and data sources, especially during migration periods.
A consistent orchestration enables organizations to leverage cloud resources with a low level of complexity. The end goal of orchestration is to make applications and services portable, able to run in any cloud without requiring software or configuration changes.
Simplify Monitoring and Management
Public clouds, local resources, and private cloud systems have their own management and monitoring tools. Using multiple tools in a heterogeneous environment often limits end-to-end visibility for operations teams and creates unnecessary complexity.
To enable cohesive management of a hybrid cloud, organizations need a single management solution they can use across infrastructure and applications running both in public clouds and on-premises.
Ensure the Environment Meets Policy and Governance Requirements
IT policy compliance includes everything from legal compliance, to industry standards, to internal organizational policies. When moving data and workloads to the cloud, maintaining compliance in a unified and consistent way is critical to reducing business risk.
By integrating governance into hybrid clouds at every stage, systems and employees can make architectural changes without violating critical requirements.
Automate your Infrastructure
IT automation is particularly important in a hybrid cloud environment. Automated management promotes self-service and reduces IT labor, reduces the risks associated with human error, simplifies policy enforcement, enables forward-looking predictive maintenance. At the end of the day, a hybrid cloud that is well automated is easier and cheaper to build, maintain, and operate.
Learn more in our guide to cloud automation
Hybrid Cloud Migration Best Practices
Here are some important best practices to consider when migrating your workloads to a hybrid cloud model.
- Create a reusable enterprise pipeline that manages the entire cycle from development to production, that is applicable for any application, meets the concerns of all stakeholders, is highly auditable, and is the only valid source of information throughout the process.
- Set up deployment automation both for local and cloud deployments. Deployment automation should be invoked non-interactively via containers or existing on-premise applications. This allows you to change the environment in stages, without affecting other stages of the development pipeline.
- Use a single source for configuration and refer to the same single source from both source and destination architecture. Do not use the same configuration multiple times, to avoid synchronization problems, and gradually reduce migration overhead.
- Do not migrate all applications at once. If you have hundreds of applications, you cannot migrate them all at once. What you need is a migration pipeline which you can apply to each application, to take it through the necessary steps until it becomes part of the cloud environment.
- Remember that each application is different. While a migration project should aim for standardization, you should recognize that each application portfolio may have a completely different migration strategy.
- Look for opportunities to utilize cloud-native architectures and interfaces, and reduce legacy technology that is not well suited to the cloud environment.
Learn more in our guide to cloud migration strategy
How Does Application Dependency Mapping Support Cloud Migration Management?
Hybrid and multi-cloud environments often require migrating workloads from one cloud to another, or between on-premise and cloud infrastructure. In any migration challenge, a major challenge is application dependencies - components like other applications, networks, or storage equipment that applications rely on, and might break or malfunction if they cannot properly access them.
When migrating applications to the cloud, it is often necessary to refactor code and address dependencies that will be absent or work differently in the cloud environment. Without a solid understanding of the current interconnections between applications, databases, and other resources, it can be very difficult to migrate to the cloud.
Application dependency mapping (ADM) is an IT management tool that shows a complete map of application dependencies - including connections, data flows, databases, and other resources. This can allow developers to consider how each dependency will behave in the migration. In many cases, it can be possible to provide a similar dependency in the cloud environment to prevent extensive code refactoring.
Application dependencies are a primary cause of failed migrations, and even if migrations do succeed, it it is difficult to extend and scale applications in the cloud without fully understanding their dependencies. ADM can be instrumental in reducing the complexity, cost, and risk of application migrations.
Learn more in the guide to application dependency mapping
Hybrid Cloud vs. Multi Cloud Management
It is not enough to have a hybrid cloud environment or migrate your applications to it, you also need a robust set of tools to manage the hybrid cloud environment on an ongoing basis. Let’s look at the nuanced differences between the technologies used to manage a hybrid cloud (typically with one private cloud and one public cloud) vs a multicloud.
Hybrid Cloud Management
Hybrid cloud management is the process by which an organization controls the deployment of private and public cloud infrastructures. Administrators can achieve this by consolidating the infrastructure of public cloud vendors and on-premise cloud resources into a single management platform, with one administration UI, typically provided by a dedicated cloud management tool.
Most cloud computing service providers offer hybrid cloud management solutions that connect their infrastructure and services with services from other providers and vendors, allowing organizations to run two or more platforms through a single user interface. In principle, this is a reasonable goal to achieve, because private and public clouds are both based on essentially the same hypervisor technology.
Hybrid cloud management tools should allow administrators to create or destroy instances, assign workloads to resources, and view performance characteristics across the hybrid cloud infrastructure.
Learn more in our guide to hybrid cloud management
Multicloud management provides a consistent workflow for managing infrastructure setup, security, connectivity, and service discovery across any cloud platform. Well-defined multi-cloud management gives you visibility and control which can significantly reduce the complexity of the multicloud environment.
Multicloud management platforms provide the following key benefits:
- Self-service capabilities—eliminate existing processes associated with IT resource delivery, allowing consumers to provision their own resources.
- Workflow automation—enable organizations to create and manage compute instances and more complex cloud configurations without manual intervention.
- Cloud analytics—analytics and monitoring allows organizations to use the best available services as needed. Metrics allow organizations to automatically switch cloud providers and move workloads between clouds to improve cost or performance.
Hybrid Cloud Platforms
Hybrid cloud is not just a methodology. In the end, the hybrid cloud runs on a technology platform. There are many platforms that can be used to create a hybrid cloud, some of them mainly focused on the private cloud aspect, and some mainly focused on integrating public cloud elements with an on-premise environment.
Below we review four of the most common used solutions:
- Microsoft Azure Stack
- VMware Cloud
- Google Anthos
OpenStack is an open source multitool software platform that allows you to build and manage public and private clouds. OpenStack was originally launched in 2010 with support from NASA. OpenStack provides an Infrastructure-as-a-Service (IaaS) solution designed to achieve simple implementation, massive scalability, and enterprise grade security.
OpenStack allows users to manage their cloud environment by deploying virtual machines and other resources, scaling them up and down, and monitoring their deployment.
OpenStack uses virtual resources to run a combination of open source tools that help organizations manage and maintain their cloud. After deploying each of these tools, OpenStack users can access specific components and APIs that provide cloud infrastructure capabilities.
OpenStack has declined in popularity over the past decade due to its enormous complexity and steep learning curve. The OpenStack community worked hard to reduce this complexity, but many organizations are still finding it easier to pursue other alternatives.
Red Hat OpenShift offers an enterprise-grade Kubernetes platform, built especially for an open hybrid cloud strategy. OpenShift provides a wide range of container technologies, including container orchestration software based on OKD, which is OpenShift’s open-source project.
Red Hat OpenShift offers security and administrative features that help manage Kubernetes on a large scale. It is available as a fully-managed cloud service deployed on many top public clouds, and also as a self-managed software that can be customized to meet organizational requirements.
Learn more in the guide to OpenShift Container Platform
Microsoft Azure Stack
Azure Stack is a portfolio of products that extend services and capabilities offered on the Microsoft Azure public cloud, to local data centers, edge locations, remote offices and other local environments. It builds, deploys, and consistently runs applications across location boundaries, giving organizations the flexibility of truly hybrid deployment.
Common use cases of Azure Stack include:
Edge and disconnected solutions
For example, remote or mobile locations with an unstable Internet connection. In these cases, Azure Stack provides full cloud infrastructure on the local device, without an internet connection.
Cloud applications with regulatory requirements
Azure Stack allows organizations to leverage cloud technology and benefits while hosting data and resources in their local data center, to meet regulations and industry standards.
Bring the cloud application model on-premises
Azure Stack enables organizations to run applications on-premises using a cloud deployment model, and easily burst capacity to the public Azure cloud if local resources are exhausted.
VMware Cloud is a service that integrates with all major public cloud providers, including AWS, Microsoft Azure, Google Cloud, IBM Cloud and Oracle Cloud, as well as 4,000 VMware Cloud Provider Partners. The promise of VMware Cloud is to allow users to manage, connect and protect any application in any cloud.
VMware Cloud integrates private clouds built using VMware technology with public cloud technology, making it possible to manage workloads in a consistent, hybrid environment, with a single operational and security model. This minimizes risk and reduces complexity for hybrid cloud projects.
A key tenet of VMware cloud is that applications, which are currently running on-premises using VMware technology, can be transported to any cloud without modification, reconfiguration, or transformation. This can simplify application modernization strategies, and reduce overall migration risk.
Learn more in our guide to VMware cloud services
Google Anthos is an open hybrid and multi-cloud application platform that helps organizations modernize existing applications, build new applications and run them anywhere in a secure manner. Anthos is built on open source technologies developed by Google—Kubernetes, Istio, and Knative—which work together to provide consistency between local and public cloud environments.
Anthos is an upgraded version of the former Google Cloud Services Platform (CSP). Unlike CSP, Anthos can manage workloads on Google's competitors: other public clouds such as AWS and Azure. While competing cloud services offer a hybrid stack, Anthos differentiates itself by supporting both hybrid cloud and multicloud (the environment most companies currently use).
Anthos accelerates the development process by providing a consistent environment for developers to create applications and deploy them to the cloud of their choice, allowing developers to focus on functionality rather than compatibility between on-premises environments and other public cloud environments.
Administrators can use Istio (the Google Service Mesh which supports both containers and virtual machines) as a secure communication channel between applications, services, and end users.
Learn more in our guide to Google Anthos
Virtual Desktop Infrastructure (VDI), Desktop as a Service (DaaS) and Hybrid Cloud
Many organizations are using virtual desktop infrastructure (VDI), enabling corporate users to run a virtualized desktop from anywhere. VDI represents a large investment for many enterprises and is a prime target for hybrid cloud strategies.
What is VDI?
Virtual Desktop Infrastructure (VDI) is a technology for desktop virtualization that runs desktop operating systems, usually Microsoft Windows or Linux, and manages it in a local data center. The virtual desktop image is delivered over the network to the endpoint device. This allows users to interact with the operating system and its applications as if they were running locally. Endpoints can be traditional PCs, thin client devices, or mobile devices.
Learn more in in these detailed guides>VDI Solutions
What is DaaS?
DaaS (Desktop as a Service) is a cloud computing product that allows businesses to serve virtual desktops to any device. A DaaS solution provides a fully managed desktop for applications that can be delivered securely over the web.
Unlike VDI, which is an on-premises solution that requires heavy maintenance, with DaaS, local IT teams do not need to buy and maintain software and local infrastructure. DaaS providers manage VDI deployment, maintenance, security, upgrades, data backup and storage. The customer is responsible for managing the application and desktop images.
DaaS is offered as a subscription service and is multi-tenant by default. In the back end, the cloud provider is offering the same VDI technology as many organizations manage in-house. The DaaS provider streams the virtual desktop to the customer's end-user device.
Learn more in our guide to Desktop as a Service
A Hybrid Cloud Strategy: Connecting VDI and DaaS
The COVID-19 epidemic and the uncertainty accompanying it has created many challenges for enterprise IT in relation to virtualized desktop services. In many cases, the entire workforce suddenly needs access to VDI. If an organization relies only on a local data center for VDI, it can be difficult to extend the infrastructure to accommodate this sudden scalability.
This is leading many organizations to consider a hybrid deployment model, which allows them to leverage their existing VDI infrastructure, but leverage cloud bursting to extend capacity to the cloud.
A hybrid VDI/DaaS model offers the following benefits:
- Can provide the necessary IT resources in a timely manner, without requiring large infrastructure projects, because the underlying infrastructure already exists.
- Leverages geography to bring end users and remote desktops closer together.
- Provides almost infinite bandwidth and networking resources, capable of handling a large number of incoming connections.
- Possible to establish secure, reliable connection from the public cloud to the local VDI infrastructure, for example to enable transfer of user data and configuration from the on-premises VDI system to the cloud.
Fully migrating VDI to the cloud increases network latency and can cause a deterioration in user experience, especially when scalability is considered. In these cases, the true value of the hybrid cloud architecture is clear. Several vendors offer hybrid architectures that combine VDI and DaaS, with easy migration of workloads between local and cloud environments. This provides:
- Simplified migration to the cloud in a short period of time without architectural changes.
- Reduced network latency due to greater use of connections between public cloud and local data centers.
- Better security through micro-segmentation, leveraging the local network configuration.
- Cross-cloud migration capability (if you need to move your VDI platform between different public clouds).
Finally, when normal IT operations resume, virtual machines hosting the application can easily be moved back to local infrastructure without the need to rebuild the entire application, reducing consumption and costs of public cloud resources.
Hybrid Cloud Management with NetApp Cloud Volumes ONTAP
NetApp Cloud Volumes ONTAP, the leading enterprise-grade storage management solution, delivers secure, proven storage management services on AWS, Azure and Google Cloud. Cloud Volumes ONTAP supports up to a capacity of 368TB, and supports various use cases such as file services, databases, DevOps or any other enterprise workload, with a strong set of features including high availability, data protection, storage efficiencies, Kubernetes integration, and more.
In addition, NetApp Cloud Manager provides the UI and APIs for management, automation and orchestration, supporting hybrid & multi-cloud architectures, and letting you treat pools of storage as one more element in your Infrastructure as Code setup.
Cloud Manager is completely API driven and is highly geared towards automating cloud operations. Cloud Volumes ONTAP and Cloud Manager deployment through infrastructure- as- code automation helps to address the DevOps challenges faced by organizations when it comes to configuring enterprise cloud storage solutions. When implementing infrastructure as code, Cloud Volumes ONTAP and Cloud Manager go hand in hand with Terraform to achieve the level of efficiency expected in large scale cloud storage deployments.
See Additional Guides on Key Hybrid Cloud
NetApp, together with several partner websites, has authored a large repository of content that can help you learn about many aspects of hybrid cloud. Check out the articles below for objective, concise reviews of key cloud storage topics.
Authored by NetApp
Hybrid cloud architectures give organizations more control, but introduce management and security complexities. Learn how to simplify hybrid cloud management.
See top articles in our hybrid cloud management guide:
Authored by NetApp
Efficient virtual desktop infrastructure (VDI) operations are critical for business continuity. Learn how VDI works, and best practices to remotely deploy workstations on demand.
See top articles in our virtual desktop infrastructure guide:
Authored by NetApp
Desktop as a Service (DaaS) vendors use virtual machines to host desktops in the cloud. Learn how to deliver cloud-based workstations the DaaS way.
See top articles in our desktop as a service guide:
Authored by Cloudian
Hybrid IT implementations merge on-premise and cloud resources. This ensures organizations can remain in control while leveraging the scalability of the cloud.
See top articles in our hybrid IT guide:
Authored by Cloudian
VMware offers a wide range of storage types, including options for traditional protocols like SAN and NAS, as well as Software Defined Storage (SDS) models.
See top articles in our VMware storage guide:
Authored by Cato
Learn about SD-WAN, a technology that is revolutionizing the way organizations manage connectivity across wide area networks (WANs).
Authored by Hysolate
Learn about virtual desktop infrastructure (VDI) solutions that can help you deliver virtualized desktops to thousands of employees at low cost.
Authored by Faddom
Learn about application dependency mapping, which provides a comprehensive map of an entire IT ecosystem, and is especially useful in cloud environments.