Data nested with application logic is slowing cloud deployments. What businesses are left with is technology aspirations clashing with workload realities.
Published April 22, 2021
Senior Editor
Nigel Treblin via Getty Images
Hype and fuss surround the cloud, with analysts pointing toward its omnipresence in future enterprise technology deployments. For some, the momentum of innovation depends on cloud-based workloads; companies without cloud will be left behind.
Yet, businesses are a long way from cloud-only deployments.
In a few decades, "probably after we're all long retired, we want to have separate data, we want to have it structured a nice way, we want to have it all in a cloud appropriately, but we are so far away from that," said Jim Whitehurst, president of IBM, speaking during a hybrid cloud panel hosted by Protocol last week.
The underlying problem is, data is directly woven into the legacy application logic, which impacts when and how businesses can move them, according to Whitehurst.
Companies are left with technology aspirations clashing with workload realities. To migrate applications to a cloud environment, businesses have to abstract data before moving a workload — and each legacy application can require different techniques to move.
The migration creates hybrid technology that will stick for years to come. Though cloud surpassed data center spending this year, businesses have a long way to go before modernizing their technology stacks. And while there is incentive to change over time, businesses are not in a rush.
"Part of why things don't shift more quickly is that ultimately the cost and effort of doing so just doesn't have a return," Yvonne Wassenaar, CEO of Puppet, during the panel.
For a lot of companies, moving data to the cloud is an afterthought, according to Eric Drobisewski, senior architect at Liberty Mutual Insurance. Unwinding the data models that support many applications is tough and usually conflicts with other business products.
To tackle the problem, Liberty Mutual is starting with small, low-risk data sources to build migration strategies and tooling without taking parts of the business down, he said during the panel.
From there, the company plans to get into bigger, more challenging data stories, giving them time to develop a composition model and determine how to execute the migrations in line with existing business priorities.
"Ultimately, our goal is to really transition out of data centers and move all of these systems and services into public cloud," Drobisewski said.
A hybrid inevitability
There are a few companies known for advanced cloud deployments, including Capital One, which closed its final data center last year. But the shift took eight years to complete.
In the midst of a migration, technology stacks default to hybrid, encompassing a portfolio of applications running across deployment models, from on-premise to the edge.
"I don't think of hybrid cloud as a choice, to be honest. It's an inevitability," Whitehurst said. "It would be almost impossible to be any material size and not consume computing resources from multiple locations."
Cloud is aspirational as many businesses aren't built for it, though they may have made progress.
What's important is creating a hybrid bridge that takes technologies core to critical business processes and ties them to innovation a business is putting into the cloud, said Drobisewski. Hybrid is a piece of this, enabling workloads across the spectrum of technology.
What's next for businesses is a slower rate of cloud adoption as organizations take stock of what they've enabled and how it fits into their long-term plans.
Post-pandemic, people want optionality to sit back and think through their three- to five-year tech strategies, according to Wassenaar.
"They don't know all the answers today," she said. "They know directionally where they want to go. And so they want to work with folks who are going to help them through that process, and not lock them into a given answer, but give them some choices across" technology stacks.