On Thursday, April 25th, we held an event on “Achieving Data Consistency in Distributed Systems” at Prostor in Novi Sad, attended by over 110+ IT professionals. The session provided practical insights into how modern distributed architectures balance performance, availability, and data accuracy.

The session was delivered by our colleague Adriano Maljković, a Head of Development with extensive experience in developing event-driven and microservice architectures. Throughout his career, Adriano has worked on building scalable, cloud-based systems focused on data reliability and security. He guided participants through complex concepts such as the CAP theorem, ACID principles, eventual consistency, and transactional patterns, using examples from real-world projects.

The presentation began with an overview of the evolution of software architectures, from traditional monolithic systems to today’s microservice-based distributed environments. Adriano demonstrated how adopting a database-per-service approach provides greater flexibility but also introduces challenges in maintaining data consistency across services.

Through practical examples, attendees saw how traditional transactional consistency concepts (ACID) are complemented or replaced by modern approaches such as eventual consistency and asynchronous communication, achieving a balance between availability and accuracy.

One of the most engaging parts of the presentation focused on event-driven architecture. Adriano explained how using events as the primary means of communication between services allows systems to be decoupled, scalable, and resilient to failures.

Key techniques demonstrated included:

• Transaction Outbox Pattern – ensuring atomicity between local changes and event emission
• Message Brokers – for reliable event storage and delivery
• Compensating Transactions – restoring system consistency after errors
• Idempotent Event Handling and Dead-Letter Queues (DLQ) – safely reprocessing failed events
• Horizontal Scaling and Stateless Services – enhancing system availability and resilience

Particular interest was drawn to the Saga pattern and conflict resolution via versioning and timestamps, enabling concurrent changes in the system to be processed without data loss.

Adriano also shared frequent mistakes engineers make when implementing distributed systems:

• Forgotten compensating transactions
• Incomplete event payloads, making conflict resolution difficult
• Lack of idempotency, preventing safe retries
• Uncontrolled behavior when using Promise.all without transactional guarantees

Through these examples, attendees gained a realistic view of how theoretical principles are applied in practice and why it is critical to design systems with resilience from the start.

To conclude the session, attendees received a carefully curated list of recommended resources, including:

• Patterns, Principles, and Practices of Domain-Driven Design
• Microservices Patterns
• Hacking the system design interview

After the presentation, attendees joined an open Q&A and networking session with Adriano to discuss topics such as data synchronization between microservices and challenges with eventual consistency in cloud environments. The atmosphere was inspiring, and many attendees expressed interest in more frequent technical meetups.

The presentation “Achieving Data Consistency in Distributed Systems” confirmed that distributed systems are complex, yet full of opportunities when designed thoughtfully. Adriano demonstrated that consistency is achievable, not through rigidity, but through flexible architecture and clearly defined communication patterns.

CYBERLAB continues to organize events that connect the community, inspire dialogue, and promote knowledge exchange among experts.

See you at the next CYBERLAB event!

We’re always happy to share our knowledge and experiences. Feel free to follow or reach out to us on LinkedIn | Facebook | Instagram