Melpomene - Fotolia
The isolated data center -- traditional, converged, hyper-converged, composable, etc. -- is quickly becoming a relic. Hybrid cloud IT is not only our present-day reality, it is also, at this point, a necessity for architecting a modern, optimized IT ecosystem.
More than half of the IT organizations the Enterprise Strategy Group (ESG) has surveyed use public cloud infrastructure services, and over three-quarters of those utilize more than one public cloud provider. IT is more distributed and disaggregated than ever before. It is essential to incorporate these realities into your buying decisions.
Augmenting the data center with public cloud services offers incredible benefits to both infrastructure flexibility and speed-of-service delivery. There is a cost, however. According to ESG research, nearly a quarter of IT organizations identify that adding the management responsibilities of public cloud resources to existing data center oversight responsibilities increases IT complexity.
In other words, building a hybrid cloud infrastructure with traditional tools and a traditional mindset makes IT more complex and is, ultimately, less efficient. As a result, IT buyers must adapt to understand that control, protection and management strategies should be designed with the hybrid cloud in mind. In fact, many enterprises are turning to hyper-converged cloud infrastructure to ease the complexities of incorporating hybrid cloud strategies into their data centers. Hyper-converged infrastructure is a modern data center architecture that frees up IT resources and enables the deployment of a cloud strategy in stages.
While alternative definitions of hybrid cloud persist, the consensus is hybrid cloud infrastructure platforms manage applications and infrastructure that span both on-premises (i.e., in the data center) and off-premises public cloud environments. This definition holds whether your organization uses an "Infrastructure Up" model -- in which you start with a data center and expand into the public cloud -- or a "Cloud Down" model -- in which you start in the cloud, and then extend by building on-premises services. To deliver maximum value, however, hybrid clouds must deliver services beyond that simple definition.
Three things to look for in a hybrid-cloud deployment
To understand what considerations are most important, ESG recently asked IT decision-makers to identify their top requirements for a hybrid cloud. Their top three most common identified responses offer a starting point that can serve as your blueprint when considering hybrid cloud offerings.
According to these respondents, hybrid clouds must:
Manage both on- and off-premises environments. This sounds obvious, but it is rarer than you might think. Multiple hybrid cloud offerings can integrate some of what the cloud has to offer into an existing on-premises infrastructure environment or offer some visibility. While valuable, they become more limited as hybrid cloud strategy requirements increase over time. Hybrid clouds must manage on- and off-premises infrastructure as equal partners with a single interface, one that is intuitive to the user. Also inherent in this requirement is the concept of multi-cloud. It is powerful to use a hybrid cloud that supports both on-premises infrastructure and the services of one public cloud provider such as AWS. It is more beneficial, however, to seek out heterogeneous infrastructure flexibility both on and off premises, with multi-cloud support.
Provide seamless movement of applications between on- and off-premises infrastructures. Still often debated is the question of whether the hybrid cloud infrastructure is a collection of some applications that run on premises and some that run in the cloud (i.e., never the two shall meet), or whether movement across environments is required. Seamless application movement's place among the top three responses suggests application movement is a valuable and necessary capability, even if your organization does not have an immediate need for it.
Also reflected in this response is the level of control required -- namely, the need to enable application movement, not just data movement. Infrastructure, data and networking are all in service of the application. A hybrid cloud deployment that helps to manage the whole application simplifies the environment and frees IT administrators from having to manage the individual components separately. This a major advantage to hyper-convergence and is one of the reasons the hyper-converged cloud has gained steam as the basis for multi-cloud and hybrid cloud deployments for some enterprises.
Be able to deploy and expand both on- and off-premises infrastructures. This requirement separates hybrid clouds that offer only visibility from those that offer actionable control. Visibility is valuable, but consolidating control allows for staffing consolidation while providing better interaction and interoperability. A hybrid cloud strategy should do more than just manage disparate infrastructure silos. On- and off-premises infrastructures should be treated as elements of a single larger whole.
A hybrid cloud strategy is not just about "adding a little cloud to the data center." It must provide a consistent wrapper across multiple heterogeneous infrastructures while offering the ability to manage at the application-workload level. When evaluating offerings, understand that a hybrid cloud deployment should create a common framework for infrastructure compatibility with common APIs for simpler application development. They should also have an interface that looks familiar and fulfills your organization's needs.
Several offerings exist that consolidate these heterogeneous, hybrid and multi-cloud application-level management capabilities. One example is VMware Cloud Foundation, which offers hybrid cloud deployment management and control across multiple on- and off-premises options, such as VMware Cloud on Dell EMC and VMware Cloud on AWS. VMware Cloud Foundation offers deployment and lifecycle management features, along with a management interface that will be familiar to data center administrators who are using the same vSphere tools and processes for automating servers, storage and networking.
Along a similar trajectory, Kubernetes-based hybrid clouds, such as Red Hat OpenShift and Google Anthos, offer application-level control for container environments across multiple on- and off-premises environments.
The recently announced Oracle partnership with Microsoft Azure adheres to these desires as well, highlighting workload movement to the cloud without needing to rearchitect the application.
Just the beginning
Though already standard practice, the hybrid cloud deployment revolution is far from over. As organizations demand greater levels of infrastructure freedom across both on- and off-premises providers, as well as more control at the app layer, new offerings will continue to emerge while existing ones evolve even further.
For today, however, when making hybrid cloud deployment architecture decisions, follow the advice of organizations that are further along on this journey. Look for hybrid cloud infrastructures -- hyper-converged cloud-based or otherwise -- that offer application-level management, movement and control across the data center and across multiple public cloud infrastructure services.