Eclipse Digital - Fotolia
The industry realization about some of the cloud's downsides has led to repatriation measures that bring essential workflows back to the data center.
Depending on the scale and size of the migration, there are a lot of moving parts to ensure success and determine just how much cloud repatriation can cost an organization.
Considerations for cloud repatriation costs
Factors for cloud repatriation costs include the following:
- Data transfer.
- Hardware requirements.
- Available personnel.
- Archiving costs.
- Associated downtime with the migration.
But, before IT teams dive into any sort of migration, organizations must pinpoint why cloud technologies aren't useful. According to Randy Armknecht, managing director at Protiviti Inc., cost and available admin skills are potential reasons for leaving the cloud.
Cost-wise, organizations may have performed a lift-and-shift migration without fully understanding that full cost efficiency in the cloud is often found through rearchitecting legacy systems, not simply moving them over. There are also many add-on features and costs that weren't anticipated that can begin to add up, Armknecht said.
The technical skills needed to operate in a cloud setup are different than an on-premises data center, and the team may not have had enough training, leaving admins to struggle with management. The skills gap, combined with a general talent shortage for cloud engineers and architects, can make it hard to hire and keep these roles in an organization.
It's also possible the move to the cloud was poorly thought out. Ninety-two percent of IT organizations that repatriated applications had originally moved them without any rearchitecture or modification. Furthermore, 68% of organizations said the decision to move to the cloud was made by a business group outside of IT, according to Scott Sinclair, senior analyst at TechTarget's Enterprise Strategy Group (ESG).
Components of cloud repatriation costs
From a practical perspective, businesses that repatriate workloads selectively choose workloads that have particular performance, uptime, maintenance or security needs or require personalized applications and on-premises hand-holding, said Hyoun Park, CEO and principal analyst at Amalgam Insights.
Going back to on premises or building on-premises infrastructure isn't easy or necessarily straightforward. Hardware purchases, data transfer, backup and recovery, security, maintenance and infrastructure downtime are among the top considerations.
"This means thinking about networking and bandwidth costs, storage, processors, archiving costs, a comprehensive physical and digital security management strategy, and the human cost of managing and maintaining these technologies," Park said.
"Financial cost alone is rarely a strong reason for cloud repatriation, unless the company can reasonably be expected to recuperate millions of dollars in annual savings over a multiyear period," he said.
In this financial calculation, organizations should account for the costs of technical administration, connectivity, business continuity, security and replacement hardware.
Prepare for a business model shift
The challenge with cloud repatriation is not so much the initial migration as it is the ongoing maintenance, governance and risk associated with the new infrastructure. The most significant cost of moving back from the cloud is most likely the time associated with planning the project followed by the additional time that may be needed to effectively scale and support the workloads on an ongoing basis, Park said.
Companies considering cloud repatriation must consider how much of the on-premises technical management is a core organizational competency. Management should then weigh the value of their time in supporting these resources compared to the value of tackling other IT issues, he said.
Armknecht said potential cloud repatriation costs include data egress charges for moving the data from the cloud back to the data center, which can be significant, as well as new hardware acquisition costs to support VM and container sprawl. Therefore, IT teams should conduct a rationalization process to determine what workloads to bring back on premises.
Workloads that face large data transfer costs are good candidates to move to an on-premises location that is closer to avoid bandwidth costs.
Admins must also revisit and refine their disaster recovery (DR) and business continuity plans for the new on-premises setup. There are important questions to consider during DR evaluation, such as if the organization will use the cloud to back up on-premises resources and whether the data center space is in a company-owned facility or elsewhere.
Depending on the organization's depreciation schedule, IT managers should determine how repatriation affects hardware acquisition and IT spending cycles.
Support cloud repatriation with tools and timing
Once organizations select which workloads to migrate back to on-premises infrastructure, IT teams must evaluate any potential migration tools and the timing of the cloud repatriation process.
Though the IT infrastructure market has continued to push for organizations to move to the cloud, migration tools can be used to move out of the cloud as well, Armknecht said.
"We've seen an increased interest from clients in converged infrastructure, which promises to provide a layer of encapsulation above the data center and cloud service providers so that workloads can seamlessly move between them," Armknecht said.
He also said that each of the major cloud service providers has also released, and will likely continue to develop, services that enable the integration of on-premises computing and storage with a cloud management console. Current options include Azure Arc, Azure Stack, AWS Outposts and Google's Anthos.
Repatriation efforts tend to not only be costly, but they can also be disruptive. According to ESG's Sinclair, the firm's data showed that 43% of organizations that returned workloads on premises incurred costs related to downtime.
This means the timing of cloud repatriation is essential, especially if the workloads are mission-critical or an essential part of business operations. How much data IT teams must move and what infrastructure requires installation affect the amount of downtime and if the organization can support those costs.