In 2026, as hybrid cloud adoption reaches maturity, the focus of technology leaders has shifted from simple infrastructure migration to deep resource rightsizing for legacy systems. Organizations currently in the active phase of cloud transformation often face a harsh reality: computing power bills significantly exceed initial forecasts, turning architectural debt into direct financial losses.
Attempting to migrate legacy monolithic architectures to a hybrid environment using a "lift-and-shift" approach—simply copying infrastructure without code modification—inevitably leads to over-provisioning and a rapid increase in total cost of ownership (TCO). To halt uncontrolled spending, engineering teams are integrating the FinOps operational model with architectural refactoring.
The 'lift-and-shift' trap: why blind legacy migration destroys cloud budgets
Legacy monolithic code and suboptimal database structures were originally designed for on-premises servers, where hardware was purchased once to handle peak loads. In cloud environments, billing occurs every second based on allocated or consumed resources. Inefficient architectural decisions, such as "heavy" database queries or a lack of horizontal scaling, require excessive computing power, which critically impacts the budget.
Beyond financial risks, legacy infrastructure poses significant security threats. According to the ENISA Threat Landscape 2025 report, digital infrastructures and services accounted for approximately 27.7% of analyzed data breaches during the reporting period. Attempts to compensate for legacy system vulnerabilities using additional cloud security tools and monitoring often only complicate the infrastructure and increase TCO.
Anatomy of FinOps 2026: three phases of cost management in hybrid environments
The FinOps Foundation methodology defines cloud financial management not as a one-time configuration, but as a continuous operational process consisting of three key phases: Inform, Optimize, and Operate.
- Inform: At this stage, achieving cost transparency is critical. Tag-based attribution is used to precisely identify which legacy modules or business functions are consuming excessive cloud computing resources.
- Optimize: Resources are adjusted to meet actual needs. This involves implementing automated management policies to shut down development and testing environments during non-working hours, as well as rightsizing—aligning allocated capacity with actual application load.
- Operate: Continuous monitoring and the distribution of financial responsibility between development teams and operational units. In hybrid environments, this means constantly balancing local resources and the public cloud to maintain efficiency.
Architecture first, cloud second: how the domain metadata model solves rightsizing
In accordance with the Azure Well-Architected and AWS Well-Architected Frameworks, cost modeling and optimization at the design stage are significantly more effective than attempting to reduce infrastructure consumption after the fact. Effective FinOps begins with refactoring the architecture of legacy systems.
A proven solution is the transition to model-driven development (domain metadata model). Replacing resource-intensive legacy code with a domain-oriented metadata model allows for a substantial reduction in server load. Instead of writing voluminous imperative code for intermediate application servers, developers describe the data structure, interface, and logic as declarative metadata.
This allows the platform to generate the necessary REST API and optimize database performance automatically. By eliminating redundant code, memory and CPU consumption are reduced, enabling the use of smaller cloud instances and avoiding over-provisioning.
Reducing TCO in practice: database optimization and eliminating redundant code
For the migration and modernization of legacy systems, especially in the public sector or critical infrastructure, it is advisable to use solutions with a minimal infrastructure footprint. An example of such a foundation is the full-stack JavaScript low-code / model-driven platform UnityBase. This is a joint development by companies within the Intecracy Group alliance, where InBase is a key developer.
Products built on the UnityBase platform, such as the Megapolis.DocNet system or Scriptum solutions, demonstrate a significant reduction in TCO due to several architectural features of the platform:
- DBMS-agnostic ORM: Interaction with databases is carried out through a unified metadata layer. The platform automatically optimizes queries and resolves excessive load issues at the DBMS level, avoiding situations where inefficient monolithic queries necessitate scaling database computing power.
- Elimination of heavy intermediate servers: By using an asynchronous non-blocking HTTP(S) server, solutions based on UnityBase require minimal resources compared to traditional monolithic applications, which significantly simplifies rightsizing in the cloud.
- Flexible storage management: Separating the database from file storage (BLOB stores) allows unstructured content to be moved to cost-effective object cloud storage (compatible with AWS S3), leaving only transactional metadata in the high-speed database.
Enterprise and Defence editions of UnityBase: security and performance for critical infrastructure
When choosing a platform for hybrid environments, it is important to consider architectural compliance with regulatory requirements. The open-source version, OpenUB (OE), under the Apache 2.0 license, has limitations regarding use in organizations associated with the Ukrainian public sector.
For high-load systems and projects with increased security requirements, the commercial Enterprise (EE) or Defence (DE) editions are officially recommended:
- Enterprise Edition (EE): Provides support for enterprise-level databases (including Oracle RAC), integration with Active Directory and commercial authentication providers, and multitenancy support.
- Defence Edition (DE): Designed for mission-critical systems. It includes authentication via public/private keys, integration with certificate authorities (CRL, OCSP), support for national encryption standards (DSTU), and integrity verification of server modules.
While model-driven development does not eliminate all migration risks, it provides a reliable technological foundation for the controlled transfer of legacy systems, allowing hybrid cloud resources to be used with maximum financial efficiency.
Comparison of legacy system migration approaches in the context of FinOps
| Comparison criteria | "Lift-and-Shift" approach | Refactoring for microservices | Model-driven transition (UnityBase) |
|---|---|---|---|
| Over-provisioning risk | High risk of over-provisioning, retention of inefficient code, and uncontrolled TCO growth in the cloud. | Reduced risk of over-provisioning, though potential additional resource costs for orchestration. | Minimal RAM/CPU consumption due to automatic DB query optimization and a lightweight runtime. |
| Impact on cloud TCO | High recurring costs due to payment for redundant computing power. | Significant initial investment and high operational costs for maintenance. | Radical TCO reduction due to the elimination of heavy intermediate servers and storage flexibility. |
| Complexity and Time-to-Market | Low initial complexity, fast migration, but problematic cost management thereafter. | High development complexity, long implementation time (Time-to-Market). | Rapid design via metadata, automatic API generation, and hybrid cloud readiness. |
FAQ
How can the FinOps framework be implemented when migrating legacy systems to a hybrid cloud?
Implementation is carried out as a continuous process through three phases: Inform (identifying costs through attribution and resource tagging), Optimize (rightsizing and automating the shutdown of non-working environments), and Operate (continuous monitoring and distribution of financial responsibility between teams).
Why does model-driven development (domain metadata model) reduce cloud infrastructure costs?
It allows for the replacement of heavy imperative code with declarative metadata. This eliminates the need for resource-intensive intermediate servers, automatically generates APIs, and optimizes database queries, minimizing the load on computing power.
What is the difference between the Enterprise and Defence editions of UnityBase for government information systems?
The Enterprise edition supports commercial DBMS (such as Oracle RAC), multitenancy, and Active Directory integration. The Defence edition is additionally expanded for critical infrastructure: it supports integration with certificate authorities (CRL, OCSP), server module integrity verification, and national encryption standards (DSTU).