8 Cloud Migration Challenges

Cloud maturity is increasing, but there are several challenges that can make it more difficult or expensive than it needs to be. The adoption pattern has also evolved, in part, due to the pandemic.  

In the beginning, cloud migration was about storage, followed by applications and compute. With all the cloud services available now, businesses can achieve new levels of efficiency and take advantage of a growing set of capabilities, though cost remains front and center. 

“Before COVID, we had to do all these pricing analyses to prove that cloud was going to be cheaper and we also talked about total cost of ownership,” says Paul Borghese, VP of cloud and infrastructure at digital transformation consulting company Publicis Sapient. “Now it’s about cost optimization. For a lot of our customers the low-hanging fruit has been moved to the cloud and now the biggest issue is expense.”  

Some of Publicis Sapient’s customers are using ServiceNow or JIRA for automated billbacks based on a billing code. Once the request has been approved and the automatic landing zones have been created, the environment is automatically scripted with approved designs for security, storage and networking. Two benefits of billbacks are that IT’s budget isn’t being taxed unnecessarily by other parts of the business and when the user is finished, they simply click a button that deletes everything and ends the billing. 

Related:Top 5 Skills Gaps Every Cloud Practitioner Needs to Close

Ragu Rajaram, global cloud consulting leader at multinational professional services network EY, says while shifting to cloud-based services will reduce costs in the medium to long term, the initial migration process has huge initial costs associated with it. 

“To save costs, many companies prefer to migrate on their own which can cause inefficiencies and end up costing more in the long run. For instance, organizations may take workloads to the wrong environment or spend valuable time remediating skyrocketing consumption cost,” says Rajaram in an email interview. “Each year, organizations spend more due to inefficiencies in cloud migration process and a substantial number of companies see their migrations delayed. This unanticipated cloud spend adds up globally in wasted spend. To avoid this waste, organizations should define clear KPIs to determine what they expect to save or spend after shifting to cloud. It helps determine whether migration is economically efficient or challenging.” 

Following are some more cloud migration challenges organizations face. 

1. Application dependencies 

Related:8 Priorities for Cloud Security in 2024

Migrating data centers to the cloud presents numerous challenges. According to Will Huber, office of the CTO at digital transformation platform provider AHEAD, the most common one is that organizations don’t understand the critical dependencies between applications or even components of a single application. 

“Understanding how apps talk to one another and what core services that are essential for them to function is important to factor into migration plans,” says Huber in an email interview. “Failure to do so can lead to costly outages, performance, and user experience degradation and possible security risks. Companies that take the time to do a thorough application dependency mapping and factor their finding into their migration planning increase their chances of success in their cloud transition.” 

2. Data security and regulatory compliance 

Cloud migration reveals new vulnerabilities and risks. Teams can leverage the cloud to protect their cloud development pipelines by using infrastructure as code (IaC) techniques, according to Richard Marcus, CISO at audit, compliance & risk management software provider AuditBoard. 

“IaC makes it easy to incorporate security controls at the code level and evaluate new infrastructure for issues using tools integrated into the development process. Most public cloud platforms have even made tools available to evaluate and monitor their development pipelines for security issues,” says Marcus in an email interview. “An emerging class of capabilities such as Cloud Security Posture Management (CSPM), and Cloud Native Application Protection Platform (CNAPP) are being used by security teams to provide real-time visibility into security misconfigurations or otherwise suspicious activity that requires investigation. Integration of these tools into the cloud management control plane, and build pipeline, gives security teams a level of visibility and control they could never achieve in an on-prem environment.” 

Related:Deciphering Cloud Signals: 2024 Cloud Trends Report

3. Old thinking patterns 

To truly harness the potential of the cloud, Jan Strnad, CTO of global travel platform provider Daytrip says you need to modernize your way of thinking.  

“It’s not that hard to provision and decommission carbon copy environments in minutes for testing, deploy[ing] many times a day, or hav[ing] a hassle-free production deployment if you utilize modern approaches like infrastructure as code, serverless, edge computing, etc. The final solution is usually simpler, not more complicated, even if it sounds like the opposite,” says Strnad in an email interview. “At Daytrip, we migrated service by service and, after the whole shift to serverless, Docker, and CDN, we were able to massively speed up the delivery, mainly thanks to CI/CD, which removed the entire release management overhead." 

4. Faulty assumptions 

Migrating to the cloud can be difficult for organizations that have historically relied on on-premises data centers, mainframe computing, and traditional release cycles. Todd A. Jacobs, executive director of the Theia Institute Think Tank, says effective cloud migrations can challenge core assumptions in key areas such as: 

“Scaling vertically, horizontally, or even regionally is easier in the cloud than in physical data centers. It is not always cheaper or more labor-efficient, though, so it's important to consider all the potential trade-offs before fully committing to cloud migration as a strategic centerpiece,” says Jacobs in an email interview. 

5. Not thinking critically enough 

Horizontal and vertical migrations both have challenges. 

Kyle Campos, chief technology and product officer at cloud management platform provider CloudBolt Software, says horizontal migrations leave the workload architecture largely untouched and are able to find a more efficient cloud provider in the same public or private cloud context. This is suddenly more interesting now given the M&A activity in the private cloud space (e.g., Broadcom) that has teams re-evaluating cost models for private clouds and potential hypervisor changes. 

“This is a complicated endeavor and must be done with a refreshed economic model and a sober evaluation of your team’s skill set and capacity to manage not just the migration itself but the day two ops of where the workload lands,” says Campos in an email interview. 

Vertical migrations are typically “lift and shift” style migrations moving from private to public cloud or, less often, from public to private. This is typically a first step to reduce the surface area of private cloud and ease any risk or toil in private cloud horizontal migrations that may be necessary. Public cloud services leveraged for this are a very stable and well-described commodity which simplifies the movement. Lastly, there’s a modernization migration, which includes not just the workload movement itself but also changing the architecture patterns of the workload to better align with cloud native patterns. 

“As a business, you need to continually evaluate many dimensions of inputs to decide what blend of techniques is right for your business at any given time. These dimensions include which workloads contribute most to differentiating capabilities in your business, how often you need to redeploy/modify the usage and scaling patterns, compliance needs, and how the economics scale,” says Campos. “Make sure you have a common management place for balancing these factors and focus teams on putting the right workloads in the right cloud, with the right economic scale factors to ensure the highest cloud ROI.”

6. Moving operational technology to the cloud  

Dustin Johnson, CTO at advanced analytics software provider Seeq, says while the initial focus of cloud endeavors was migrating IT workloads like storage and backup to the cloud, the current challenge is moving complex and critical operational technology (OT) systems, integral to manufacturing processes to the cloud.  

“These systems are not only intricate and deeply embedded in operations, but also demand specialized knowledge from OT administrators, who may not be trained in cloud services,” says Johnson in an email interview. “Furthermore, [keeping] OT data secure and compliant with various standards presents a significant hurdle, reinforcing the divide between cloud IT workloads and manufacturing operations.” 

To bridge the gap between IT and OT and derive value from cloud adoption, manufacturers should leverage the cloud's scalability and advanced analytics capabilities, says Johnson. For instance, utilizing SaaS platforms that routinely undergo SOC 2 assessments can offer a secure and compliant environment for sensitive OT data, enabling businesses to adapt to changing needs, analyze vast amounts of data for better decision-making, and empower their workforce with actionable insights. 

7. Legacy system knowledge and technical debt 

Ram Palaniappan, chief technology officer at IT services management company TEKsystems, says most applications going through modernization have multiple years of business and process knowledge built in with little to no documentation so reverse engineering is difficult. 

“Adding to the complexity, many of the functionalities and business logic associated with legacy systems and technical debt may need to be depreciated [or] discontinued,” says Palaniappan. “This business decision and sign-off is a major challenge. To overcome this, it will be crucial to have legacy system experts on the team who can help traverse this inherent system knowledge efficiently.” 

8. Talent and change management 

Most cloud skills need significant upskilling from on-premises tools and systems.  

“Addressing this challenge will play a key role in migration efforts and can look as simple as instilling skill training to employees to further their growth,” says Palaniappan. 

Two related areas, also people-centric, are change management and user adoption. 

“Any new system comes with a user comparison to an older system. To overcome this, there needs to be a strong change management program to drive user adoption.”