According to Gartner, by 2021 over 75% of midsize and large organizations will have adopted a multi-cloud or hybrid cloud IT strategy. With this widespread acceptance of these deployment models comes the question — what is the perfect fit for your organization? Multi-cloud or hybrid? Christopher G. Chelliah, SVP for customer strategy, business development and insight for Japan & Asia Pacific at Oracle, shares why the choice should be a combination of both.
Q: For many organizations, multi-cloud is the go-to strategy when selecting a deployment approach, why?
A: Many organizations are adopting a multi-cloud strategy for three key reasons. Firstly, cloud economics. Choosing different cloud service providers allows customers to reduce costs and optimize the overall IT expenditure. Secondly, the capabilities it brings to the table. As enterprises deploy different workloads, they are likely to discover differences in cloud offerings. A multi-cloud strategy gives companies the flexibility to use the best possible cloud for each workload. Thirdly, availability. Distributing applications among multiple cloud service providers (CSPs) offers greater availability and resiliency to cloud failures.
Q: So if multi-cloud offers such benefits is there a need to adopt hybrid too?
A: Today’s business environment is evolving rapidly, with digital transformation happening at a tremendous pace. It is not a ‘question of either/or’ for hybrid and multi-cloud. It is about creating the right platform for the organization to gain a competitive edge with a mix of both strategies.
Multi-cloud has certainly become commonplace for many organizations. At the most basic level, nearly all companies use multiple clouds – for example, a webmail server alongside an online productivity tool. It has come about as most organizations have a ‘hyperscaler’ of choice as their preferred provider for key cloud workloads and then use other clouds that were needed to satisfy cost, workload, or availability requirements, as discussed.
Hybrid cloud, also known as distributed cloud, is about using different cloud deployment types across on-premises, private and public cloud- and can include multi-cloud in some cases- to achieve a goal. The choice is typically based on the need some organizations have to run most of their critical workloads are on-premises for regulatory and performance-related reasons, for instance, companies in healthcare, financial services, and the public sector. But it can also be driven by latency requirements or the desire for control.
Before I move on it is also worth noting that we are seeing that many organizations have adopted different services from different vendors, thinking that they are operating in a multi-cloud environment. Really they are just operating multiple clouds rather than multi-cloud. Not understanding the nuances and not approaching it strategically typically leaves a lot on the table when it comes to a CIO managing enterprise IT.
Here at Oracle, we have shown our commitment to multi-cloud over the last few years. We have a cloud interoperability partnership with Microsoft. It enables users to virtually seamlessly connect Azure services, like Analytics and AI, to Oracle Cloud services, like Autonomous Database. By enabling customers to run one part of a workload within Azure and another part of the same workload within the Oracle Cloud, the partnership delivers a highly optimized, best-of-both-clouds experience, and things like a unified identity and access management, unified single sign-on experience, and a collaborative support model.
Q: So to get the maximum agility, flexibility, and cost control should a company use a multi-cloud architecture, for instance with cloud storage from one provider and a SaaS solution from a different one?
A: In practice, it boils down to what works best for the workloads and the governance framework of those workloads-some workloads will only run on a specific cloud environment. For example, Oracle has the only cloud infrastructure platform that offers Exadata as a cloud service. So if a company is running its mission-critical workloads on Exadata on-premises and wants to match like for like in the cloud, Oracle Cloud Infrastructure (OCI) will be the best solution. Additionally, organizations look to spread capacity and risks across multiple cloud providers.
Take Zoom for an example. It runs the majority of its video conferences on OCI and another hyperscaler. A complete multi-cloud strategy means that an organization can gain the best price-performance for a specific workload with the ability to shop across the hyperscalers. That said, not all hyperscalers are thrilled at giving the customers the choice of going multi-cloud.
Q: What about sensitive workloads, should they be kept on-premise while more common business processes are migrated to the cloud. Is this hybrid approach considered a less risky and smart move?
A: Customers are still constrained as to which workloads should stay on-premises and which ones can be migrated to the cloud. To help ease the burden of choice, our strategy is to offer customers a virtually seamless experience, a virtually seamless set of services, a the virtually seamless rate card, a seamless set of certifications, regardless of whether the workload is deployed on the public cloud or in a hybrid infrastructure on-premises.
To help with that last year Oracle launched its Dedicated Region Cloud@Customer. This is a solution that helps bridge the gap between cloud and on-premises environments. Unlike other on-premises cloud offerings, which deliver only a very small subset of the services available in their public cloud regions, Oracle makes more than 50 services available on-premises so that enterprises can use Oracle’s cloud services wherever they need them — in the cloud or on-premises.
Additionally, Oracle has a cloud VMware offering that allows customers to leave some portion of their estate on-premises and live migrate or burst others to the public cloud. And they can do this without having to upgrade or modify people skills, tools, processes, and policies – and, importantly, with complete control. This means customers can get to hybrid cloud in a matter of hours without any change and the associated risk.
Q: Ultimately, what would be the ideal approach? Is it a multi-hybrid-cloud approach?
A: Multi-cloud helps customers spread their dependence and risk across different cloud providers, helping them to enhance efficiencies for workloads that require a precise service only available in a specific cloud. This approach also gives customers the bargaining power for cost efficiencies across the hyperscalers.
For highly regulated or security-focused businesses with workloads needing to meet demanding latency and data residency requirements, a hybrid cloud then delivers the same benefits and experience as the public cloud.
With organizations needing to access both these sets of benefits, most organizations will need a holistic digital transformation plan with a strategy for both hybrid and multi-cloud.
