Hybrid Cloud Computing is a strategy that utilizes a private cloud (or on-premises data center) with one or more public cloud offerings that are connected via public or private networks. The key capability that links the two environments, is consistent operations that allows the public cloud to act as an extension of a private or on premises environment, with compatible management process and tools.
Organizations often adopt a hybrid cloud strategy to maintain proprietary or protected information in local data centers, while also enjoying the scale and pay-per-use benefits of public cloud infrastructure. For example, legal may wish to store and process all its data on-premises, while marketing wants to use public cloud services to interact with employees and customers, keeping web traffic off of the local corporate network. Hybrid cloud computing commonly utilizes consistent management operations across environments.
7 Reasons VMware Cloud Foundation is the Premier Hybrid Cloud Solution: An IT Executive Guide
Most professionals cite agility and cost as the primary advantages of hybrid cloud computing. Here are the top benefits of adopting this approach to computing:
- IT Flexibility. Since cloud provider capabilities vary, Hybrid cloud lets organizations put workloads where they can take advantage of provider features or to meet exacting regulatory requirements. For example, sensitive information could reside on-premises, while web server workloads run on the public cloud, keeping traffic from clogging the local network.
- Security and Compliance. Although most cloud providers offer physical security for their servers, many providers can further isolate an organization’s cloud infrastructure by utilizing dedicated servers and private networks to connect to on-premises gear. Additionally, public cloud providers can help an organization meet specific needs imposed by governments, for example by offering a presence in Europe for GDPR-protected data.
- Predictability. Although cloud providers can meet the demands for most of today’s workloads, organizations know the capability of their own on-premises infrastructure. Those applications where latency is an issue might be better suited for on-premises hardware.
- Pay-as-you-go flexibility. By utilizing ‘cloudbursting’ to extend IT capabilities into the public cloud when demand spikes, organizations need only pay for the additional capacity when needed rather than purchasing enough equipment to meet spikes in demand.
- Development Sandboxes. Development teams can experiment with new tools and operating systems available on public cloud platforms. Public cloud resources can be used to develop and test new applications or used for application modernization without impacting users or production data. The ease of spinning up new environments can also help foster innovations, such as piloting new products or trying out new applications.
There are several hybrid cloud computing models. A hybrid cloud deployment could combine a public cloud with on-premises infrastructure, such as legacy mainframe-based applications. Or, it could combine public cloud with a private cloud, which could either be on-premises or hosted on rented infrastructure. A hybrid cloud deployment could also combine multiple public clouds, private clouds, and on-premises legacy hardware.
Hybrid cloud deployments all share some common characteristics:
- A single management tool or platform that eliminates the need to manage each cloud element separately.
- Connectivity either via public internet or private network connections. Usually, virtual private networks (VPN) will connect the elements of a hybrid cloud environment.
- Control plane consistency across clouds so organizations have ‘a single version of the truth’ at all time.
- Application program interfaces (API) that enable the sharing of information between programs and cloud platforms.
There are several considerations when determining whether to adopt a hybrid cloud computing strategy. First, organizations must determine whether a hybrid approach is necessary, specifically whether there is a need to have coordination and communication between workloads that reside in public cloud and on private cloud or on-premises servers. Many organizations my opt to migrate everything to public cloud providers or keep on-premises workloads and data logically isolated from public cloud workloads.
Next, organizations must determine which cloud computing platforms will comprise their hybrid IT environment. For example if there will be a need for certain workloads or data to reside in a specific geographic area, they should choose a cloud provider with a presence there. Although the major platforms such as Amazon AWS, Google Cloud Platform, and Microsoft Azure all offer similar features, some workloads may be better suited to a particular platform’s unique features. For example, an organization might opt to host Windows or Hyper-V workloads on Azure while hosting the bulk of their VMware VMs on AWS. Regardless of which platforms are chosen, organizations should seek to ensure they are not locked-in to a particular vendor and have the capability to move workloads and data as needs change or to find the most attractive price for a given feature set.
Once providers are chosen, organizations must determine which workload goes where, focusing on performance, pricing, accessibility, compliance, and the needs of the given applications workload. The effort to migrate legacy mainframe workloads to a cloud platform may not be worth the time or money involved, and so those workloads probably should remain on-premises, along with any data that is subject to governance or regulatory mandates that demand they be secured on-premises.
All hybrid cloud workloads need to be managed and orchestrated, hopefully by a single platform or set of tools. Once workload locations are determined organizations must choose a cloud operating system (OS) framework such as VMware Cloud to manage their overall cloud environment.
Finally, organizations need to understand that public cloud providers operated on a shared-responsibility security model. Although the physical security offered by cloud providers is formidable, IT professionals must consider that every connection is suspect, and should adopt a zero-trust approach to securing cloud workloads. This must ensure that every single transaction presents the proper credentials to help prevent breaches or data loss from occurring. Organizations should choose a security approach that provides visibility into both public and private aspects of a hybrid cloud deployment to simplify security management across the enterprise.
Virtually every organization with on-premises infrastructure will eventually adopt a hybrid cloud computing strategy as new services and functionality such as modern applications built around microservices and APIs continue to grow in popularity. Those organizations that have clearly defined goals and a solid understanding of application and security needs will find their hybrid cloud journey greatly simplified.
As in many areas of IT, expertise in hybrid cloud computing is in demand. 90 percent of organizations report cloud skill shortages, according to 451 Research 2019 Trends in Cloud Transformation report, which can big a major obstacle to adoption of hybrid cloud computing solutions.
Many organization will gain the most from a hybrid cloud environment by modernizing their existing workloads to better take advantage of cloud functionality. Although this is the desired end state for many businesses it involves a major commitment to rearchitecting applications.
Coordinating security can be difficult in a shared responsibility model. Regulatory compliance and visibility into security challenges are the two biggest security issues according to a Cloud Research Partners 2018 Cloud Security Report.
Outages do occur, so organizations should have the ability to fail-over and fail-back to and from cloud infrastructure to ensure uninterrupted delivery of applications and services to end users.