markrubens - Fotolia

Multi-cloud adoption is the future, so be prepared

The move towards multi-cloud is imminent for most enterprises, according to IDC. Find out what this means for organizations and how they will need to adapt.

Multi-cloud is the new normal for enterprises -- a reality that presents a number of challenges for IT teams trying to get a handle on their disparate cloud resources.

By 2021, more than 90% of enterprises will rely on multi-cloud to meet their evolving IT needs, according to IDC. This shift is driven by the evolution of the on-premises data center and the increasing sophistication of cloud and edge computing. Multi-cloud adoption is also being pushed by developers and data analysts in line-of-business teams that move to the cloud for service functionality, analytics capabilities, geographic distribution and more, according to Mary Johnston Turner, research vice president at IDC, who discussed the topic in a recent presentation.

The term "multi-cloud" has been around for years but still lacks a single definition. IDC takes an expansive view on the term, counting any mix of IaaS, PaaS, SaaS, private cloud or on-premises dedicated hardware with cloud attributes as a multi-cloud deployment. But those pieces are rarely interconnected, which creates a problem for administrators.

Enterprises must then decide how to manage, standardize, govern, optimize and use all these different platforms, Johnston Turner said in an interview. With so many moving parts, it can be an intimidating task to adopt a multi-cloud architecture -- if the organization isn't well-prepared for the move.

Enterprises' next step in multi-cloud adoption

Multi-cloud architectures add complexity. Each platform has its own set of rules for operations and management, which makes it harder to cross-train staff and give IT teams a holistic view of a company's cloud and on-premises assets.

Mary Johnston Turner IDC FutureScapesMary Johnston Turner

"Most organizations are pretty clear that they do not want broad silos operating independently in perpetuity," Johnston Turner said.

Enterprises want to be able to optimize these different assets within their own constraints around security, cost and performance. That's why IDC predicts 70% of enterprises will deploy unified VMs, Kubernetes and multi-cloud management processes by 2022 to facilitate standardized governance and to provide a single view across environments.

"You're just always going to have to do the talking to the lower level systems," she said. "But increasingly, the level of [automated extraction] -- the policy and controls that you can put on top of those controllers -- is really what matters."

That's why nearly every major cloud vendor or third party has moved toward a strategy that offers a common control plane across cloud and on-premises environments. Multi-cloud adoption is also partly fueling the adoption of containers, which will become the core fabric for many of these environments -- allowing for developers and vendors to standardize and automate CI/CD tool chains, according to IDC.

Multi-cloud success requires cultural change

Multi-cloud adoption represents change, which isn't always welcome. "The issue [with multi-cloud adoption] is that humans change slowly, and we have humans in control [who] are making these decisions," said Johnston Turner.

Organizations will need time and planning to institute common standards and policies, which will have an impact on IT. Just because developers think multi-cloud is the right choice, it doesn't mean decision-makers in an organization will immediately agree. So while most of the early investment in cloud has been siloed, it's up to IT teams to bridge those gaps and show how it will benefit the business as a whole.

"We recognize that we need the portability, scalability, consistent security, compliance, performance and cost control," Johnston Turner said. "But getting everyone on the same page is going to take some work and a lot of [that work] is cultural and organizational."

Dig Deeper on Cloud deployment and architecture

Data Center