Organizations want to make the most of their cloud investments, but not all applications are cloud-ready. Here’s what to consider when migrating to the cloud.
With 90% of organizations claiming success with their multicloud strategy, according to a 2022 HashiCorp survey, it’s time to shift our attention from whether an enterprise should embrace multicloud to how to effectively do so. As with any relatively new corner of technology, the problem and opportunity is people. That is, today there’s a dearth of multicloud talent, but there are ways around this relatively short-term issue.
Benefits to multicloud adoption
Though multicloud used to be more aspirational than real, today’s enterprises are multicloud by default. The reasons aren’t hard to guess: A multicloud strategy enables an enterprise to pick the best cloud service for their particular needs. So they might look to Google Cloud for their machine learning services, AWS for serverless and Microsoft to seamlessly connect to their existing Windows investments.
SEE: Research: Managing multicloud in the enterprise; benefits, barriers, and most popular cloud platforms (TechRepublic Premium)
Though companies will cite a desire to evade lock-in as a key reason for multicloud, in practice this is hardly the most compelling reason, given that it’s precisely each cloud’s particular strengths that enterprises should embrace to drive innovation. For years AWS would — rightly, in my view — pillory the idea that smart cloud strategy involved sticking to the most basic primitives to ensure cross-cloud compatibility. By contrast, today’s enterprise is happy to pick a particular cloud provider to match their application needs, which will differ from application to application.
Yes, along the way organizations establish independence from their vendors. But you needn’t to evenly distribute services across clouds to avoid undue concentration in any one cloud. Given how quickly the clouds roll out new, innovative services, there will always be a rich variety of services to help organizations meet their application needs.
Challenges associated with multicloud adoption
Whatever the benefits, stiff challenges exist. Foremost among these is people. For years it has been difficult to find sufficient talent to staff corporate cloud projects, and multicloud exacerbates this. There simply aren’t many people that are proficient in more than one cloud provider.
In fact, as revealed in the HashiCorp survey, a talent shortage is the number one inhibitor to multicloud success (Figure A).
It’s unlikely things will remain this way, however, given corporate demands and personal ambition. Beyond simply padding one’s resume, multicloud awareness offers real opportunity for an IT professional’s employer.
“More than just helping you absorb new information faster, understanding the strengths and tradeoffs of different cloud providers can help you make the best choice of services and architectures for new projects,” said Forrest Brazeal, head of content for Google Cloud.
So, yes, there’s a talent shortage in the cloud, and it’s even worse for enterprises hoping to succeed with multicloud. More clouds, with fewer people that understand how to translate between the diverse tools and workflows of particular clouds, means that succeeding with multicloud is far from automatic. Still, organizations are finding clever ways to circumvent this issue.
How to implement multicloud adoption
As the HashiCorp survey found, “the top approach to solving this conundrum is to standardize on a common cloud operating model — a set of common tools and automated workflows chosen and enforced by platform teams leveraging the most skilled personnel.” I’ve written before about similar platform-oriented approaches, and how they give developers a paved path to building applications.
Enterprises shouldn’t want their developers having to learn the nuances of how AWS or Azure implement security. Every minute they spend combing through those docs is a minute they’re not building more innovative applications. Platform teams can take the lead on figuring out such nuances and give developers “a standard, pre-approved environment in which the effort to create an app from an idea is minimal, [developers can] focus on innovation not plumbing,” as Weaveworks CEO Alexis Richardson has explained.
Which is why such platform teams top the list in Hashicorp’s survey (Figure B).
Such platform teams take on the heavy lifting of figuring out which cloud offers the best tools for a particular workload. It’s on these teams that the multicloud unicorns will exist. They’ll know, for example, whether Google’s GKE or AWS Lambda will be the best fit for the company and how to shave costs and boost performance with new chip architectures.
Part of this platform approach is to embrace tooling that is already multicloud. A variety of SaaS providers, like Confluent for data streaming, offer out-of-the-box multicloud architecture. Such services don’t complete a multicloud platform journey, but they offer a significant head start. For example, though a database multicloud service won’t take care of architecting your application tier for multicloud, it does remove an otherwise difficult task of managing multicloud for the data tier. Platform teams can effectively manage multicloud complexity across the different tiers of their workloads.
Over time, multicloud expertise will grow within organizations. Employees will move between companies, and even within a particular organization, skills upgrades happen incrementally on an as-needed basis. Platform teams can provide a much-needed bridge to this time when we’re all multicloud by both choice and necessity.
Disclosure: I work for MongoDB but the views expressed herein are mine.