ANALYSIS OF THE APPLICATION OF THE JAVA-ORIENTED STACK (SPRING FRAMEWORK, POSTGRESQL, REDIS) FOR BUILDING FAULT-TOLERANT ENTERPRISE DIGITAL SOLUTIONS

ABSTRACT

The article examines the use of a Java-oriented technology stack, including Spring Framework, PostgreSQL, and Redis, for building fault-tolerant enterprise digital solutions. The principles of microservice architecture, mechanisms of replication and automatic failover, as well as the role of caching and in-memory data processing in improving performance are analyzed. Practical case studies of Uber, Robinhood, and Graybar are presented, demonstrating the scalability and resilience of systems under high load conditions. The study concludes that this technology stack is applicable for implementing digital solutions with a predictable level of availability and the ability to scale linearly.

АННОТАЦИЯ

В статье рассматривается применение Java-ориентированного стека технологий, включающего Spring Framework, PostgreSQL и Redis, для построения отказоустойчивых корпоративных цифровых решений. Анализируются принципы микросервисной архитектуры, механизмы репликации и автоматического переключения при отказе, а также роль кэширования и обработки данных в оперативной памяти в повышении производительности. Приводятся практические кейсы компаний Uber, Robinhood и Graybar, демонстрирующие масштабируемость и устойчивость систем при высоких нагрузках. Сделан вывод о применимости данного стека для реализации цифровых решений с предсказуемым уровнем доступности и возможностью линейного масштабирования.

Keywords: Spring Framework, Spring Boot, PostgreSQL, Redis, fault tolerance, microservices, scalability.

Ключевые слова: Spring Framework, Spring Boot, PostgreSQL, Redis, отказоустойчивость, микросервисы, масштабируемость.

Introduction

In the light of extensive digitalization of business procedures and growing requirements for continuous operation of information systems, fault tolerance has turned into a standard feature of corporate solutions. Contemporary digital platforms must not only be outstandingly performant and scalable but fault-tolerant in sudden breakdowns – both on network infrastructure and on internal services levels. Thus, the choice of technology stack with the capability to ensure system resilience has gained special significance. The Java-based stack consisting of Spring Framework, PostgreSQL, and Redis is popular among industry due to it being mature, highly configurable, and suitable in order to facilitate fault-resistant architecture.

The purpose of this article is to examine the applicability of the Spring Framework, PostgreSQL, and Redis stack in the design and implementation of fault-tolerant corporate digital solutions.

Main part. Overview of the technology stack components

The choice of technology stack is of vital significance in enterprise fault-tolerant solution development [1]. Of interest is the stack that consists of Spring Framework, PostgreSQL, and Redis – an experienced, adaptable, and business-proven combination of components commonly used in modern information systems.

Spring Framework, being the foundation for server-side Java application development, achieves Inversion of Control (IoC) and Dependency Injection (DI) as a component container that achieves loose coupling between modules as well as support for handling object lifecycles. The most widely used style of application development in Spring is declarative definition of components and their assembly via annotations or Java config, making it possible to develop extensible and flexible designs with ease. Usage of Spring Boot accelerates application development due to auto-configuration and bundled servlet containers (for instance, Jetty or Tomcat) and also on top of that makes containerization and CI/CD easier.

To be able to provide a better picture of the strengths of the Spring framework, table 1 provides a structured overview of its core modules and key technical aspects.

Table 1.

Core spring framework modules and their technical characteristics [2]

According to the State of Spring 2024 survey [3], in which more than 1,400 developers participated, the most used modules of the framework are Spring Security (76 % of participants), Spring Data (73 %), and Spring WebMVC (70 %). Spring Session adoption is also on the rise and is at 36 % as of 2024, an increase of six percentage points over the previous year.

PostgreSQL is an industrial-strength object-relational database management system offering strict compliance to ACID semantics, full compliance with the SQL standard, and extensibility. It supports both physical (streaming) and logical replication, synchronous and asynchronous modes with tunable replication lag, providing deterministical control over consistency and latency among nodes. The key features of PostgreSQL are summarized in table 2.

Table 2.

Key capabilities of PostgreSQL and their technical characteristics [4]

Summarizing the key capabilities of PostgreSQL highlights its potential for building robust enterprise systems. This conclusion is reinforced by industry survey results. According to the State of PostgreSQL 2024 report [5], the ecosystem shows clear signs of maturity and a growing professional community. The share of highly experienced users has increased significantly, with 21 % of respondents having more than 15 years of PostgreSQL experience, while newcomers with less than one year of experience account for only 4,1 %. An increasing number of developers use PostgreSQL for both professional and personal projects, with this figure rising to 60% – a 20-percentage-point increase compared to the previous year. The vast majority (86,9 %) report using PostgreSQL as the primary database in production systems. PostgreSQL 16 leads in production environments (61 %), whereas PostgreSQL 17 has become the standard for development environments (95,9 %). Another significant trend in 2024 is the increasing use of artificial intelligence solutions: 55,3 % of developers note that they incorporate AI into their routines (an increase on 36,9% in 2023), with ChatGPT being the most utilized solution mentioned by 72 % of adopters.

Redis is a low latency, high performance in-memory key-value data store that is ideal for real-time applications. It is typically used for pub/sub messaging, distributed locks with the Redlock algorithm, session store, and data caching. Redis Cluster and Redis Sentinel provide support for high availability through monitoring, automatic failover, and built-in sharding support. Integration with the Spring environment is offered by Spring Data Redis for both the reactive and imperative APIs as well as for cache annotations. The main Redis features and their role in enterprise systems are summarized in table 3.

Table 3.

Key Redis capabilities and their use in enterprise systems [6]

Recent studies confirm the broad adoption of Redis in professional software development. With a slightly below 23 % adoption rate among professional developers worldwide, Redis is among the most utilized key-value and cache solutions, the Stack Overflow Developer Survey 2024 [7] reports. Its uses are much more than cache alone: Redis is used extensively to cache sessions, pub/sub messaging, and microservice coordination, which indicates how it is used in building responsive and fault-tolerant enterprise software.

Collectively, Spring Framework, PostgreSQL, and Redis form a well-rounded enterprise tech stack. Spring enables modular application development, PostgreSQL offers data consistency and good scalability, and Redis offers low-latency access for real-time consumption. Combined, their union enables fault tolerance, horizontal scaling, and contemporary workflows like containerization and CI/CD.

Ensuring fault tolerance: architectural approaches

Fault-tolerant enterprise solution implementation requires a unified approach that cuts across all tiers of the system – from business logic to data storage and network infrastructure. In a Spring Framework, PostgreSQL, and Redis technology stack, system robustness is achieved through an interplay of microservice architecture, self-recovery of services through self-healing, data replication, and load balancing (fig. 1).

Figure 1. Fault-tolerant architecture based on Spring Framework, Redis, and PostgreSQL

As seen in figure 1, the fault-tolerant architecture incorporates three primary components: Spring Framework, Redis, and PostgreSQL, each supporting one another in redundancy at their level. Spring Framework forms the center, providing service isolation and resilience architectures, while the microservice method allows for individual deployment and module separation so that the system still runs even if a service fails. Resilience4j enables circuit breaker and retry mechanisms to mitigate cascading failures, and Spring Boot Actuator with Kubernetes health checks supports automated diagnostics and container restarts, creating a self-healing architecture consistent with modern backend practices and focused on reducing operational risks [8].

Redis complements the stack by providing in-memory caching and pub/sub communication between services. The Cache Aside pattern accelerates API response times while reducing the load on the database. Redis Sentinel offers automatic failover and informs clients upon topology change. Horizontal scalability is enabled by Redis Cluster through sharding and replication, distributing load and making data available even in the event of single node failure.

At the data storage level, PostgreSQL ensures reliability by synchronous and asynchronous replication and by mechanisms of logical replication for selective synchronization of one or more tables. Patroni together with distributed consensus stores (e.g., etcd or Consul) supports automatic master to standby role switching without the risk of split-brain scenarios. This ensures that downtime is minimized and business processes are guaranteed to run uninterrupted. Extensions such as Citus and the ability to perform read operations from replicas (read scaling) contribute to horizontal scalability, which is especially critical for high-load platforms, including those in financial and healthcare domains [9].

Thus, the integration of Spring Framework, Redis, and PostgreSQL provides a multilayer fault-tolerant architecture in which each of the components – from business logic and interservice communication to data storage – contributes to service continuity in its own way. Not only is single point of failure resistance ensured by this technology stack, but system scalability, automatic recovery, and responsiveness to dynamic loads are also guaranteed, making this technology stack one of the best choices for the development of modern digital platforms.

Practical implementation cases of the Java-oriented technology stack

The use of the Spring Framework, PostgreSQL, and Redis technology stack in corporate environments is confirmed by a number of scalable and fault-tolerant solutions implemented by leading companies. These systems demonstrate consistent behavior under high load, within distributed infrastructures, and under strict SLAs, particularly in the domains of e-commerce, document management, and CRM systems.

As a first example, the experience of Uber is illustrative, where a Java-oriented stack using Spring Boot and Redis forms the basis of a scalable microservice architecture. The internal JFX platform is built on top of Spring Boot and standardizes service development, providing a unified approach to configuration and API routing [10]. The comprehensive caching solution CacheFront for Docstore is responsible for reading and writing client data and is integrated with the caching layer implemented with Redis. This architecture separates the cache from the storage engine and allows each layer to be scaled independently, and also includes a mechanism for invalidating stale entries (fig. 2).

Figure 2. CacheFront design

This solution minimizes the need for vertical and/or horizontal scaling to support low-latency read requests, reduces resource allocation at the database core level, lowers P50 and P99 latencies, and stabilizes read latency spikes during microbursts.

Another example demonstrating the effectiveness of relational solutions is the experience of Robinhood, where PostgreSQL became the foundation of a scalable brokerage platform architecture. As traffic grew – from approximately 100,000 requests per second in December 2019 to 750,000 requests per second by June 2020 – the company decided to migrate from a single database instance to a sharded configuration [11]. At the application level, a routing layer was introduced to distribute incoming API requests across shards and an aggregation layer to merge results from multiple shards, while transactional events are processed through Kafka with filtering by users belonging to the respective shards (fig.3).

Figure 3. High-level architecture of Robinhood’s sharded system with routing and cache layers

Each shard has its own database and cache so that requests could be processed separately and failures would be limited to a single segment. This approach successfully reduced the load on individual nodes, improved system fault tolerance in general, and minimized the «blast radius» of any subsequent outages. The move to the 10-shard configuration provided them with near-linear scalability and enabled the platform to support sustained operational growth without compromising response times to levels that would be unacceptable for trading applications.

An illustrative example of the use of Spring Framework in a corporate environment is the experience of the American company Graybar, one of the largest distributors of electrical products, actively developing its own e-commerce channel. Facing the low reliability of its legacy e-commerce site and the ongoing costs associated with IBM Commerce support, the company entered into a partnership with TDK Technologies to migrate the outdated system to a microservice architecture [12]. As part of the project, approximately twenty REST microservices were developed using Spring Boot, which improved the speed and resilience of the online platform. In addition, caching was implemented between the frontend and the database, accelerating repeated requests and reducing server load. It is confirmed that the project achieved about 90 % code coverage and 90 % mutation coverage, and was completed in five months instead of planned six. As a result of the modernization, the speed of the site increased significantly, and fault tolerance and stability of the system were improved, allowing the company to deliver unintermittent operation of the online selling channel with growing volumes of orders.

Briefly, the tested case studies demonstrate that the Java-focused technology stack, comprised of Spring Framework, PostgreSQL, and Redis, is entirely sufficient for high-scale and fault-tolerant enterprise-level digital products. Real-world deployments at Uber, Robinhood, and Graybar demonstrate how the combined use of these technologies makes system behavior predictable under high concurrency, near-linearly scaleable, and measurably reducing response latency under load. The Spring Boot technology ensures uniform microservice development as well as maintenance, PostgreSQL ensures robust transactional consistency and stable data retention, and Redis ensures quick access to frequently accessed data as well as efficient cache expiration. The presence of automated testing and continuous testing methodologies further adds strength to these solutions, reducing operational risk and delivering stable performance at peak loads [13]. These features in combination facilitate improved availability of service, reduced failure impact, and improved adherence to demanding SLAs – impacts that are critical to e-commerce, banking services, and other data-intensive business environments.

Conclusion

The study that has been carried out has shown that the Spring Framework, PostgreSQL, and Redis stack is an ideal platform to utilize while designing corporate digital solutions with high fault tolerance and scalability requirements. All these pieces contribute to the overall resilience of the system: Spring provides microservice modularity and seamless integration with resilience patterns, PostgreSQL retains transactional consistency and strong data storage through replication and auto-failover support, and Redis relieves database load and accelerates access to critical information with in-memory processing. The complementarity of the technologies enables solutions that would be functional even in the event of single service or node failure.

Practical case studies and empirical performance metrics confirm the applicability of this stack. An additional advantage is the ability to scale without significant architectural complexity or increased maintenance costs. The results obtained can be used to develop guidelines for technology stack selection in digital transformation projects and for designing systems with a predictable level of fault tolerance.

References:

1. Bolgov S. Creating infrastructure for scalable fintech solutions: technical and organizational aspects // ISJ Theoretical & Applied Science. 2024. Vol. 140. № 12. P. 354-358. DOI: 10.15863/TAS.2024.12.140.43 EDN: ZRBTAF
2. Dimitrijević N., Zdravković N., Bogdanovi M., Mesterovic A. Advanced Security Mechanisms in the Spring Framework: JWT, OAuth, LDAP and Keycloak // Proceedings of the 14th International Conference on Business Information Security (BISEC 2023). 2024. P. 64-70.
3. State of Spring Survey 2024 Results / Spring [Electronic resource]. – Access mode: https://spring.io/blog/2024/06/03/state-of-spring-survey-2024-results (date of application: 16.09.2025)
4. Salunke S. V., Ouda A. A performance benchmark for the postgresql and mysql databases // Future Internet. 2024. Vol. 16. №. 10. P. 382. DOI: 10.3390/fi16100382 EDN: XWZWQP
5. State of PostgreSQL 2024 / S3 Amazon AWS [Electronic resource]. – Access mode: https://s3.amazonaws.com/assets.timescale.com/state-of-postgresql/Report_State_of_PostgreSQL_2024.pdf (date of application: 16.09.2025)
6. Almeida D., Lopes M., Saraiva L., Abbasi M., Martins P., Silva J., Váz P Performance comparison of redis, memcached, mysql, and postgresql: A study on key-value and relational databases // 2023 Second International Conference On Smart Technologies For Smart Nation (SmartTechCon). IEEE, 2023. P. 902-907.
7. Stack Overflow Developer Survey 2024 / Stack over flow [Electronic resource]. – Access mode: https://survey.stackoverflow.co/2024/technology (date of application: 18.09.2025)
8. Kovalenko A. Evolution of backend architecture in payment platforms: strategies for enhancing fault tolerance and efficiency in the economic context // Cold Science. 2025. №15. P. 13-21. EDN: NAOYMW
9. Garifullin R. The use of modern web technologies for ensuring compatibility and interoperability in the context of medical information platforms // International Journal of Humanities and Natural Sciences. 2025. Vol. 1-3(100). P. 99-103.
10. How Uber Serves Over 40 Million Reads Per Second from Online Storage Using an Integrated Cache / Uber [Electronic resource]. – Access mode: https://www.uber.com/en-AT/blog/how-uber-serves-over-40-million-reads-per-second-using-an-integrated-cache (date of application: 19.09.2025)
11. How we scaled Robinhood’s brokerage system for greater reliability / Robinhood [Electronic resource]. – Access mode: https://newsroom.aboutrobinhood.com/how-we-scaled-robinhoods-brokerage-system-for-greater-reliability/ (date of application: 20.09.2025)
12. Spring Boot Java Microservices / TDK Technologies [Electronic resource]. – Access mode: https://www.tdktech.com/case-studies/spring-boot-java-microservices (date of application: 21.09.2025)
13. Drogunova Y. API test automation with Java/TestNG as a means to reduce operational risks and costs in digital products // Eurasian union of scientists. series: technical, physical and mathematical sciences. 2025. Vol. 1. №3(128). P. 16-20.