.png){kind=icon}
While we all get excited by new technology, some of the most amazing tech stories are based on the duration—as much as the innovation—of the technology. The IBM mainframe was predicted to die in the 1990s, yet it persisted, and still its revenues are growing even today. Similarly, COBOL, which some thought wouldn’t survive beyond the 1960s, is still in widespread use today. Even the infamous prediction that the iPhone would fail turned out to be incorrect.
In today’s large enterprise, the typical enterprise infrastructure therefore remains understandably complex and diverse. It might include a mainframe, various Windows-based servers, Linux systems, a few proprietary UNIX machines, and a variety of typically Windows desktop devices.
And now, we have the cloud. Whether private or public, whether singular or hybrid, whichever of the myriad vendor instances, you’ll be hard-pressed to find an organization of any size that doesn’t rely on the cloud as a key platform. Cloud computing has also risen steadily and significantly in the last decade and a half, from outlier innovation to fully mainstream. It doesn’t replace everything; instead, it provides an additional infrastructure option. Interestingly, 76 percent of organizations worldwide use a multi-cloud. Cloud—if we can refer to it as a singular entity—introduces an additional layer of choice and complexity.
Elevated Ambitions
Why cloud? Well, it promised us a few enticing benefits—
- Flexibility and Scale: Cloud is ordinarily a service you commission based on need, so you can expand and reduce according to need. It ensures you can grow to the highest requirement the business may need—even if that is unknown.
- Cost Efficiency: As a result of only consuming what you need, so the theory holds, costs are therefore better managed. Gone are the days of vast sums being invested to own or lease great vistas of IT equipment, along with sourcing operators, to run the business.
- A Focus on What Matters: For many organizations—even technology ones—the data center required a complex and expensive level of technical management and operational administration. Specialist IT and administrative staff were needed to keep the "engine room" running. This is not a core function, nor a cheap one, while also taking budget and resources away that could be innovating business value.
- Ease of Use: The conceptual explanation of the cloud as simply being "someone else’s computer" implied a straightforward, simple process for migrating to it. After all, the platform, application sets, and data storage models should basically be the same.
Fogged Up
The previous section outlined the projected, anticipated benefits of a typical cloud deployment. This is perfectly true in a million cases worldwide. However, we’re also aware of quite the reverse. There are some fundamental risks, too.
It is costing more than we expected. Recent overall cloud revenues have continued an upward trend, but as this article suggests, some of that has been driven by the excitement around AI. After all, the cloud giants are also investing giant-sized sums in becoming leading AI platforms, too. Furthermore, the operational realities of clouds have seen a growing disappointment in the overall cost of service, further exacerbated by punitive egress fees, as explained in this TechCrunch article.
Cloud migration can be complicated. The simple applications and use cases of the cloud, such as desktop software, email, ERP, and CRM systems? Well, we’ve got the t-shirt. What remains as potential cloud migration journeys are the harder-to-disentangle and harder-to-replicate systems that remain on-premise because, among other things, no one has dared to try to move them, and—critically—they are too important to break. Modernization of core systems that includes a platform shift is a highly specialized and incredibly complex undertaking.
Given this, there can never be a blanket approach to presume everything is fit for a cloud odyssey; there are further checks and balances required.
To Cloud or Not To Cloud
That is the question. Some commentators assert that the cloud is a panacea for all IT needs, but it isn’t that simple. Further intelligence is required to determine what makes the cut as a candidate for the cloud. In a modernization context, for example, determining what can go into the cloud "unchanged" needs careful consideration. Some technology in the source environment—let’s say it's an AS/400 or a proprietary mainframe environment today—does not always have a ready equivalent in the cloud. So, things need to be converted—or rewired—accordingly.
Elsewhere, while results have proven that much of an IBM mainframe application can transfer over largely intact in a so-called "lift-and-shift" approach, it might not be strategically preferable just to pick up a somewhat outdated system, saddled with technical debt, with few trained experts, and hope that putting it in the cloud is going to magically make everything better.
All of which often gives rise to an "and, not or" approach, where cloud is introduced alongside other platforms and systems, with IT adopting a mixture of approaches that most sensibly fit a range of requirements. Optimizing what works and extending capabilities through a flexible cloud approach is becoming a popular—and pragmatic—approach.
Regardless, the question must be asked: What’s our strategy? What are we planning to do—and why?
Know Before You Go
Any good travel book will sagely advise on appropriate information to know when arranging a trip. Local currency, language, travel routes, customs, landmarks, accommodation, transport, and so on. In short, you need a plan.
Major IT programs are analogous and need the same level of planning. Except this time the journey is being taken by the existing business applications. The destination, depending on the strategy, may be a new version of themselves.
It is therefore imperative to consider the following—
- Will this system work in its new home, unchanged? If not, what elements would need attention?
- What are the most important functions of this system that must be preserved?
- If we take this system out of its current locale, what dependencies need to be resolved? Will what gets left behind continue to function?
- Are there elements of the current system (e.g., the language, data layer, or user interface) that require upgrading? How can we approach this?
- How do we prove the new, changed system works?
- If we change strategy or want to use more than one solution, how do we manage this?
- What facilities and means of gathering insight can we use to help us with the planning of this sort of project?
There are others—including how much it will all cost—but these questions alone should stop you dead in your tracks if you have no means of answering them. Technology is the remedy here: A major technology transformation program must use appropriate technology to assist in establishing the necessary insight, planning, and execution. For each of the above questions, forensic tools to assess and transform business applications are critical to the project’s success.
Conclusion
Periodically, but relentlessly, the IT landscape changes. Hardware replacement cycles evolve, and new technologies are always around the corner. The current era of near-ubiquitous cloud adoption did not unify the IT ecosystem; rather, it has added another layer of choice and complexity. Now, hybrid IT infrastructures, combining on-premise and cloud-based solutions, are hugely popular. And it is this modern infrastructure into which more traditional, high-value application stacks may need to be moved.
A journey towards the cloud is less about the cloud and more about managing a process of change. Achieving that with the right cost efficiencies and operational certainty requires an array of talent, best practice, and—critically—specialist technology. Using forensic insight will mitigate the many risks involved and support a more informed and successful journey.
Because, after all, before you set off to somewhere, you should know how you’re going to get there.
