Cloud spending is on track to surpass $1 trillion by 2024, due partly to pandemic-driven changes to business operations, an event that dramatically accelerated cloud adoption timelines across industries. High-value motivators for consideration when companies move technology operations to a public cloud include critical revenue-generating core line of business applications, back office, support, dev/test, and business continuity and disaster recovery (BC/DR) elements. Anticipated outcomes are myriad: improved experiences for customers, employees, and suppliers/partners, increased development agility that delivers faster time to market, and an (almost mythical) gain in operational efficiencies across organizations. These results help leaders achieve the perpetual annual budgetary goal of “do more with less.”
Migrating applications to the cloud is already driving broad business practices from an operational and capital perspective. These are nascent, as operations are still realizing a host of new best practices around cloud engagement: money, risk, timeframes, resources, and cultural considerations. The C-suite recognizes the potentially massive value proposition; it is not hard to understand why there is so much pressure from management to transfer databases, data, applications, and IT operations to the cloud.
According to new research from O’Reilly, nearly all organizations agree that public clouds are vital for continued growth, indicated by the 90% of respondents that are already using some version of a public cloud, including hybrid or multi-cloud. Of those using the cloud, 48% reported they plan to migrate 50% or more of their applications in the coming year, while 20% plan to migrate all their applications—numbers that are shocking given the sheer amount of decades-old legacy applications still running on-prem. To meet these expectations, companies require enormous changes in their foundational view of the future of IT. Still, the mandate and expectations are crystal clear: move out of the data center, move away from legacy architectures, and move into the public cloud in some form as quickly as possible.
As a highly visible corporate IT initiative, cloud migration is an oft-feared endeavor for CIOs because of the high failure rate—and failures usually mean someone’s job, or at least a roadblock to promotion. According to Gartner, 83% of data migration projects break down or exceed budgets and schedules (failure). The Bloor Group finds that more than 80% of migration projects run over time and budget. This recent article reports that for each successful migration, six to 10 fail but according to experienced industry professionals, this metric is questionable. Migration efforts can prove highly complex technically and politically, often touching almost every aspect of the business and its people.
True innovation at the enterprise level is a complex parallel process. Success requires a comprehensive strategic understanding of the challenges and pitfalls of the different approaches and players involved, with real committed buy-in from all stakeholders.
One of the primary reasons a failure state occurs is that organizations neglect (or are unable) to strategically evaluate the entirety of the environment before finalizing their migration approach. Teams are often pressured into action prematurely and migration projects advance in an ad hoc manner. While this creates a large cloud footprint, it can deter organizations from fully realizing the promised benefits of the cloud due to unforeseen and excessive costs and time. Ad hoc processes often lead to boosted cloud operational costs and sometimes to significantly increased complexity from having to maintain two environments: legacy on-prem and a fragmented cloud environment. Many “hybrid” cloud environments today are really the residuals of failed attempts at migrating mission-critical applications to the public cloud. A more holistic approach to migration can limit capital overruns and costly delays.
Not all workloads are created equal. Organizations must carefully audit applications to fully understand the requirements needed and the actual capabilities (not just marketing numbers) available in the cloud before moving them. While some applications are categorized as “low-hanging fruit” with the expectation that they garner quick wins in the cloud, larger and more complex applications (that leverage legacy monolithic databases like Oracle or MS SQL Server) may require refactoring for a successful migration. That’s almost always a show-stopper for companies that don’t wish to rewrite applications or move off their traditional database platforms. The time, cost, and risk involved are simply too great for most mere mortal customer experience officers (CXOs), and these mission-critical applications are vital to the survival of an organization. While significant benefits are achievable by migrating high-value apps, an outage or slow service interruption can significantly impact business operations, directly influencing revenue and profitability, potentially causing a host of other headaches. Social media fallout, NPS scores, degraded patient care, customer turnover, and loss of credibility are some problems that may arise. Risks are high as today’s environment extends the visibility of both success and failure equally, both internally and externally.
Successfully migrating mission-critical apps—the “money makers” or the “crown jewels”—may require special services in addition to the regular cloud-native infrastructure to meet demanding requirements for performance. Problems typically arise when trying to run mission-critical or line-of-business applications in an environment that does not fully match the resilience or performance of the on-prem production environment, triggering degraded or unavailable features or functionality, and causing unhappy customers and co-workers.
Cloud architects and chief experience officers leading the charge should begin their cloud journey by migrating less important applications or functions. This way they can start taking advantage of cloud benefits without exposing themselves to excessive risks. These include applications that are not customer-facing, nor are they in the critical path of internal workflows for analytics, reporting, or the like.
Migration Mission: Possible
As mentioned, organizations should take a judicious and cautious approach when determining which infrastructure will best serve mission-critical workloads. Inventorying the application portfolio is a significant step since application footprint and performance requirements are radically different. It’s critical to determine which applications might require refactoring or redesign ahead of migration and which will probably need special services on top of cloud-native infrastructure to meet demanding requirements for performance and resilience to keep existing feature/functionality at current (or at least acceptable) levels.
Mission-critical workloads drive concerns about data performance (IOPS, throughput, and latency) and options should be thoroughly evaluated. Sometimes applications do not work as well in the cloud as they did on-prem. In this case, ‘slow’ is the new ‘down’ and worst-case scenarios include slower performance in addition to the dreaded outage of a company’s services. Apps need to be fast, and reliable, period.
Before committing to a migration strategy, it’s absolutely crucial to understand that most on-prem applications were not architected to take advantage of cloud APIs and design elasticity; often they were purpose-built to run on bespoke hardware, storage, and networking. It makes sense that these workloads are better suited for on-prem or private cloud infrastructure, but proper design considerations can help deter cloud-native capabilities and limitations. Remember, hyperscalar architectures are built to accommodate the largest common footprints possible, not to easily hold the 10 to 20% of corporate applications that have exceptional requirements around performance, security, availability, or reliability. The public clouds need some help there.
Architecting the Right Solution
There are many variables to ensuring a solid foundation for a successful migration, including identifying the right teams (skillset and culture) for the job, and identifying the proper technology, software, and services. In addition to the right combination of people, skills, and resources is clear visibility regarding the business value and criticality of all applications. Make sure you have both a business and a technical champion who have the appropriate authority and motivation to ensure the proper time and resources are allocated to the initiative to grant a good chance for success—ideally, they also understand that even an excellent plan can’t be perfect and build in allowances for project alteration or even rollback if an unknown pops up and creates a hard blocker.
Armed with a deep understanding of the application’s cloud readiness and a sound approach to migration, CXOs must assess the costs of running the application on-prem versus in the cloud. If the numbers do not justify a lift and shift and a replatform is not viable, it is time to introduce a data platform to support the critical workloads that require higher performance, stability, and availability than is available with cloud-native, which assure that the user experience in the cloud will prove at least as good (or even better) as on-prem. A high-performance cloud data platform facilitates the migration of the entire enterprise data ecosystem to the cloud, including those systems with extremely high data gravity.
The huge benefits of the cloud are undeniable: flexibility, scalability, automation, and agility are just a few. But business leaders must also be aware of the potential risks throughout all phases of migration. The complexity of existing infrastructure, vendor monopoly, data loss, latency (slowness), wasted money (overprovisioning)—all of these add to the challenges of cloud migration. It can prove daunting, but companies can mitigate risks and ultimately enjoy the numerous benefits of moving applications to the cloud with the proper strategies and support in place. But planning is essential, or the migration project can result in a colossal failure.
Accelerated digital transformation triggered by COVID-19 is certain to persist far beyond 2022, and companies must adapt now. Cloud strategies are proven to not just help organizations move faster and reduce costs in many cases, their agility, elasticity, and global accessibility also enable faster innovation and the integration of newly emerging disruptive technologies. But to capture a piece of the trillion-dollar prize, companies require a clear view of the provisions and risks of cloud adoption when choosing the proper technology platforms to ensure success.