Moving to the cloud too fast is slowing the pace of IT modernization and causing a lot of wasted work, according to a new survey.
In a December 2019 study, research from IDG found that 41% of IT directors delayed or abandoned some of their 2019 IT modernization initiatives, often due to competing priorities or lack of a clear strategy.
The State of IT Modernization 2020 report also found that defining a successful cloud strategy requires a lot of trial and error. Sixty-four percent of organizations that have made progress with IT modernization initiatives have switched from an all-private or all-public cloud strategy to a hybrid approach. IDG conducted the survey for Insight Enterprises.
Other findings of the survey of 200 IT directors include:
- 84% of organizations moved select workloads from a public cloud to an alternative cloud or non-cloud location
- Database, web apps, dev/test, and data analytics are the workloads most likely to have been brought back in-house
John O’Shaughnessy, an infrastructure architect and senior cloud consultant at Insight Enterprises, said companies sometimes have to bring cloud projects back to on-premises networks due to unexpectedly high cloud costs, security concerns, and other factors because of inadequate due diligence in determining the right workload-to-platform alignment.
SEE: Power checklist: Local email server-to-cloud migration (TechRepublic Premium)
He said that the key to avoiding this back and forth is defining business requirements first before moving anything to the cloud.
“By taking that time up front, then the money you invest to migrate to a public cloud infrastructure will be well spent, and you’ll start to see lower costs and higher efficiencies,” he said.
Some workloads aren’t a good fit for the cloud, and making due diligence a priority will make it easier to figure that out.
Another crucial step is defining ownership of these decisions.
“If we’re going to have certain things on premise and some on the public cloud, you have to know who owns deciding what is where and defining the decision criteria,” he said.
O’Shaughnessy said that the new business environment dictated by COVID-19 is speeding up the transition to the cloud because people are seeing the weaknesses of doing things the old way.
“You can’t just call up someone to walk down the hall and reboot the machine anymore,” he said.
Another key to success with cloud projects is updating procedures and governance before moving any workloads.
“If you don’t have it figured out ahead of time, people will be very disconnected as they try to solve problems and they will find out that their use of the cloud is not very efficient,” he said.
O’Shaughnessy said the initial sales pitch from public cloud providers–deployments are fast and easy–was very persuasive.
“Lots of companies saw the benefits of moving to the cloud, but they didn’t see the work required to get there,” he said.
O’Shaughnessy starts cloud conversations with clients with a workload transformation assessment to define business requirements and understand how the work is being done. This can increase the chances that a cloud environment is the right size to handle the work and the right place to do it.
“If you move 20% of your applications to a public cloud, but don’t move the database they’re connected to, that creates a lot of communication back and forth which then becomes an unanticipated expense,” he said.
O’Shaughnessy shared one example of this problem in which a client spent 12 months of budget in four months because of unexpected cloud expenses.
Infrastructure automation is growing
The survey also found that adoption of software-defined networking (31%), moving applications to the cloud for later modernization (30%), and upgrading on-premises application infrastructure (30%) are the most likely operating model modernization projects to be completed.
O’Shaughnessy said the key to success with software-defined networks is ensuring repeatability with good documentation.
“If you’re doing this in a scripted, repeatable way, you’ll be less afraid of causing problems,” he said.
IDG conducted the survey in December 2019. Survey respondents were required to be employed in an IT, data, and/or security related job function at the director or above at an organization with 2,500 employees or more.