In the context of digital fragmentation and stricter regulatory requirements, the geography of IT infrastructure is no longer a technical detail. Today, server placement is an element of sustainability, legal security, and customer accessibility strategies. In an interview with Michael Holopov, CEO of Russian IT infrastructure company ATLEX, we discuss why more and more companies are abandoning the "monolithic" approach in favor of hybrid architectures between Russia and Europe.
Context — what has changed?
Anna Nikulina: Michael, five years ago, the question of where a company's servers were physically located was mainly of concern to IT specialists — in terms of ping and rental costs. Today, it is discussed at the board of directors level. What has changed?
Michael Holopov: Indeed, the issue of server placement has shifted from a technical one to a strategic one — it now directly affects business stability. Three key shifts have contributed to this.
Firstly, regulatory pressure. In recent years, dozens of requirements have come into force, ranging from the storage of personal data within the Russian Federation to the classification of critical information infrastructure. Violation of these regulations can result not only in fines, but also in the suspension of operations.
Secondly, the instability of global connections. Even if a company has no legal restrictions, it faces real risks: delays in equipment deliveries, disruptions in international communication channels, and difficulties in paying for foreign services. All this makes dependence on a single region — especially a foreign one — extremely vulnerable.
Thirdly, changing expectations of customers and partners. Businesses increasingly have to prove that they can operate without disruption, even in the face of external shocks. And the geographical distribution of infrastructure is becoming part of this "reliability roadmap".
Today, server placement is not about "where is cheaper", but about "where is safer and more stable". And this choice directly affects reputation, legal stability, and even the cost of attracting investment.
Trends — what does the market see in 2026?
Previously, companies chose either a completely local infrastructure or hosted everything in AWS clouds or European data centers, but today things are more complicated. What IT resource hosting models currently dominate the Russian market?
We can talk about the final end of the era of "monolithic" architectures. Today, almost no one asks themselves the question "Russia or Europe?" as an alternative. Instead, business thinks in terms of modules: different components of the system are placed where it makes sense from the point of view of regulation, productivity, and risks.
Hybrid schemes are already becoming the main model for companies operating both within Russia and abroad. For example, banks and government agencies naturally keep everything within the country, which is understandable. But even they are increasingly creating isolated test or analytical environments abroad in order to use modern AI tools without restrictions.
But e-commerce, digital agencies, and SaaS developers—the vast majority of them have switched to dual-circuit architecture:
- The core of the system — databases containing personal data, user accounts, and financial modules — remains in Russian data centers.
- External components — landing pages, marketing services, APIs for international partners — are located in Central Europe, most often in the Czech Republic, Germany, or Finland.
Why is this the case? Because it allows us to comply with Russian legislation while maintaining low latency for European audiences. In addition, such schemes provide flexibility: if one circuit is temporarily unavailable, the second continues to operate normally.
Interestingly, even companies that previously localized completely are now cautiously testing hybrid solutions — especially where computing power is needed for AI or working with open global APIs. Complete isolation is beginning to be perceived as a technological risk.
Reasons — why is business going hybrid?
The hybrid model sounds logical, but it is clearly more complicated to operate and more expensive than having everything in one place. What exactly makes companies incur these costs? What risks are they mitigating in this way?
Indeed, hybrid solutions require more sophisticated architecture and additional costs for synchronization, security, and management. But companies are willing to accept this because the cost of risk associated with concentration in a single region has become higher than the cost of infrastructure complexity.
I will analyze it in three dimensions.
First, there is the legal aspect.
Federal Law No. 152, Central Bank requirements for IT resilience, and new Ministry of Digital Development rules regarding critical information infrastructure all make it impossible to store Russian personal data or financial transactions outside the country. At the same time, if a company works with international clients, it cannot ignore the GDPR or local EU laws. A hybrid approach is a way to comply with both regulatory regimes simultaneously without compromising either jurisdiction.
Secondly — technical.
Even with ideal communication channels, the physical distance between, say, Moscow and Frankfurt causes a delay of 30–50 ms. This is not critical for a website, but it is a problem for real-time services, video conferencing, or microservice interactions. A hybrid approach allows the architecture to be decomposed so that the "heavy" logic remains closer to the user, while sensitive data remains in a legally secure zone.
Thirdly, business risks.
Imagine a company that has completely transferred its infrastructure to Europe: it loses some of its customers in the CIS due to speed and trust issues. And if it leaves everything in Russia, it may face restrictions when working with foreign partners or marketing platforms. A hybrid approach preserves both markets and reduces dependence on a single political or economic scenario.
In essence, this is not about luxury, but about a new standard of digital flexibility. Yes, it is more complicated. But in conditions where even a cable between continents can become a point of failure, such investments are justified.
Practice — how does it work in reality?
Theoretically, a hybrid sounds reasonable, but in practice, many questions arise: How can data be synchronized between jurisdictions without violating the law? How can fault tolerance be ensured? How can such a distributed system be managed without chaos?
This is one of the most common questions faced by companies that are just beginning the transition. And yes, without a well-thought-out architecture, a hybrid can indeed turn into a "swamp of connections". But over the past two years, established patterns have emerged that allow such systems to be built transparently and securely.
First, the principle of strict data segregation. Critically important data — especially personal, financial, or medical data — never leaves Russian jurisdiction. Instead of copying, anonymized or aggregated data is sent abroad: metrics, analytics, logs without identifiers. This allows, for example, AI models to be run in the Czech Republic without exporting the source data.
Secondly, the connection between nodes is established via dedicated L2/L3 tunnels, not via the public internet. This ensures both security and predictable latency. Traffic is encrypted at both the application and infrastructure levels, so that even if the channel is compromised, the content remains inaccessible.
Thirdly, each node is made autonomous. This means that if the connection to the European data center is temporarily interrupted, the Russian part continues to operate normally, and vice versa. This is achieved through local caches, replication of only non-critical components, and a clear division of responsibilities between microservices.
When it comes to management, the "infrastructure as code" (IaC) approach is increasingly being used. Configurations for all regions are stored in a single repository, but tagged by jurisdiction. During deployment, the system automatically applies only those rules that are permitted in a specific zone. This eliminates the human factor and ensures auditability.
It is important to understand that modern hybrid infrastructure is not "two servers in different countries", but a single system with geographically distributed but logically connected modules. And the more clearly the boundaries of each module are defined, the more reliably the entire structure works.
Outlook — where is the market headed?
The hybrid model is gaining momentum, but technology and regulations are not standing still. How do you think the geography of IT infrastructure will develop over the next 2-3 years? Will hybrid become the norm, or will new architectural paradigms emerge?
Hybrids are not a temporary compromise, but a new sustainable norm, and I am confident that this trend will only grow stronger in the coming years. What's more, we will see it evolve in three key areas.
Firstly, standardization.
Today, every company is forced to design hybrid architecture from scratch. In the coming years, ready-made templates will appear — especially within sovereign cloud platforms, which will initially support multi-regional deployment, taking into account all legal restrictions. This will lower the entry threshold and make hybrid accessible even to medium-sized businesses.
Secondly, the growing role of local datacenter operators.
International hyperscalers have moved on, and regional players have filled the market. But companies are no longer just looking for a "rack space" — they expect partnerships: assistance with design, reservation, and even legal expertise on data placement. Operators who can offer not only capacity but also infrastructure consulting will gain a competitive advantage.
Thirdly, automation of management.
The future belongs to systems that track changes in legislation themselves and adapt their architecture to new requirements. Solutions are already emerging where data storage policies are linked to geographic tags, and traffic is dynamically redirected depending on the status of the channel. In two to three years, such approaches will become widespread.
However, it is important to understand that technology does not replace strategic thinking. No machine can decide where to place the client API and where to place the financial base. This remains the responsibility of the IT director and company management. And the sooner a business begins to view geography as part of its digital DNA, the more sustainable it will be tomorrow.
The geography of IT infrastructure is no longer a question of location, but part of a strategy for business survival in an unstable world. Those who understand this today will be the leaders of tomorrow.
Comments