How to reduce IT cost of ownership… optimizing physical infrastructure
In the era of cloud-first and AI-first approaches, some companies still have some or all of their operations on-premise on physical infrastructure. This in and of itself is not a problem.
Moreover, it can be a good step from the point of view of ensuring data privacy or safe operation of LLM models (AI large language models). Moreover, many companies combine pure “hardware” with the principles of hybrid or private cloud.
All these options and situations only show that physical infrastructure will not disappear from business IT. Therefore, it is necessary to find the right way to its optimized operation. This is not only reflected in finances, but also in business continuity and operational resilience.
Energy as an overlooked cost of your own IT
A data center or server room is not a set of technical equipment. It is purely an environment, a space that is supposed to create ideal conditions for technical equipment.
This definition clearly fits the environment and technical elements into the status of a long-term operational commitment, not just ordinary equipment. This is because servers and their environment require an initial investment, but the financial burden of such a solution does not end there.
For proper operation, electricity, cooling, internet connection and maintenance and management are needed. Day after day, hour after hour, minute after minute. Because most companies cannot afford to have their IT infrastructure hiccup.
How we commonly view physical IT
This is based on the idea that was already mentioned in this article: we bought the servers, so we already have them, and now we will just write them off.
Stop. That’s not exactly how it works. Servers and their operating environment require constant service and care, as well as a readiness to solve problems. These are often not non-technical, so you need not only IT guys, but hardware experts to operate them.
And on top of that, the initial investment is often excessive, because the company needs to take into account fluctuations or growth. And adapting the physical infrastructure is not as easy and simple as it might seem.
The most common blind spots in server rooms and among own servers
We come to the least popular part of any analysis. It’s time to look at the pitfalls that many companies – even experienced ones – fall into when it comes to physical infrastructure.
One of the problems is the feeling that a small in-house server room will actually create negligible costs. That it is much more profitable than throwing equipment into an external data center.
This may be true, but only if the company confirms it with analyses. If it relies only on a feeling, it will inevitably overpay. Sometimes the conditions for on-premise operation may be truly favorable, but this may not be the case, especially when we take into account all the operating and one-time costs for healthy operation.
Another problematic view is the neglect of the operating environment. Cooling and backup power are often selected on the basis of “so that it works”, “so that it doesn’t get mentioned”, not “so that it makes sense”.
Over time, the company will pay for this attitude, because something breaks, the power goes out or some internal problem occurs. Then the corrective measures are quite painful because the system was not built on a good foundation.
Another problem is operating different types of systems in the same way. Measures that need to be incorporated for critical systems may not be needed for the operation of ordinary non-critical systems. Still other measures need to be considered for the operation of sensitive systems.
Each of these types poses specific risks and needs to be adapted to them already when planning operations.
Impacts on business operations and management
The key word in this section is control, and not just financial.
As soon as the physical infrastructure in a company is not perceived as a strategic iron in the fire, then the company loses control on several fronts.
Of course, it does not optimize financially, because it is simply assumed that the “baremetal” will eat something. But active work on optimizing operations is possible at any time. Moreover, the physical infrastructure does not have to be the only part of the IT infrastructure, so it is possible to optimize costs even more.
It also happens that the company does not pay enough attention to operating conditions and maintenance. Servers do run. But these miraculous boxes are quite susceptible to their environment and for healthy and long-term operation they need a specific environment and regular care from technicians. Otherwise, for example, they will overheat or become expensive vacuum cleaners. Then it is very difficult to control the operation of the systems on which the entire company runs.
And there is also a dependence on the given location and its limits. This is likely to be less of an issue when using a data center, but for example, with your own server rooms you may encounter connectivity, power or cooling limits. In that case, your infrastructure is more in control, and you simply won’t let it go beyond certain limits.
So control is not just about finances, but also about the company’s ability to operate at its best and get the most out of the investment in physical infrastructure.
The right perspective: the server room as an operational asset
So how do you adjust your attitude so that servers are a great servant, not a bad master?
The key is to start from the needs of your company. Privacy is a really high priority for many companies, so it makes sense that they invest in it with on-premise physical solutions. But there are also systems that don’t need such a luxury. So you need to find the right place for physical infrastructure in the entire architecture of your company’s operation.
And while we’re talking about company architecture, don’t forget that just as a company cannot do without production or logistics (add your own key activities), it also cannot do without a suitable IT infrastructure today.
The right perspective comes from knowing the needs of your company and perceiving IT infrastructure as a strategic decision. Then it is possible to find an intersection that effectively covers the operation.
Classification of systems according to criticality
Now that we have the right perspective, let’s divide the company’s operations into 3 categories that need to be evaluated when planning the architecture. They will also help you find out in which areas it will be appropriate to choose physical infrastructure and where you can reduce costs using other technologies. Because yes, reducing the costs of physical infrastructure may involve moving workloads (company activities) to another environment.
Standard systems do not have special requirements for privacy and resilience. These are applications for which you will not mind if their operation goes down. The goal of deploying them on infrastructure will be to minimize costs.
Sensitive systems work with customer data, for example. Here there is great pressure to comply with legislation, especially GDPR and NIS2, so you will select the infrastructure with a view to minimizing the risk of data leakage or misuse.
Critical systems are the pillars of your operation and you need their operation to never go down. Therefore, their deployment will be accompanied by minimal downtime.
Of course, this is a simplification of the issue, and quite often you will encounter a combination of the last two types. However, it is clear that you must take into account the needs of the given system category and also the needs and responsibilities of your company. At this intersection, you will then find out what role the physical infrastructure will play in the entire company ecosystem.
Questions decision makers should ask themselves
This is a very unpopular part because it forces managers to focus on technical information. But if you see this information as strategic information rather than technical jargon, you may be surprised at how beneficial it can be for optimizing operations.
- Do we know what the TCO of our physical infrastructure is?
- Which systems really need to run locally and why?
- Do we have reserves where it really makes sense?
- Are we actively thinking about IT operations, or are we just “keeping it alive”?
Summary: calm and rational management
Physical infrastructure will never leave the corporate environment. Over time, it will simply find its place where it is truly beneficial to the company. But this requires the right approach, a change in attitude and conscious management.
Optimizing IT infrastructure should not be about companies getting rid of physical infrastructure at all costs. It is about aligning needs and strategic direction. This state then ensures that business IT is under control, not that you are under pressure.
