International Journal of Science, Engineering and TechnologyAuthors: Pramani Kota, Nallireddy Anu, Buya Lekha, Vasudev Sharma
Abstract: The increasing adoption of Java microservices deployed on cloud native platforms has intensified operational complexity, exposing limitations in traditional monitoring approaches that rely on reactive alerts and isolated performance metrics. This study addresses the growing challenge of detecting latent and cascading failures in distributed Java microservices before they materialize into service disruptions. The primary objective is to design and evaluate a cloud native observability and telemetry framework capable of enabling proactive failure detection through correlated runtime signals rather than post incident diagnosis. The research adopts a mixed methodological approach that combines system architecture design, controlled failure injection experiments, and quantitative analysis of telemetry patterns collected from JVM level, application level, and platform level instrumentation. A unified telemetry pipeline is proposed that integrates metrics, logs, distributed traces, and JVM behavioral signals into a context aware analytical model. Empirical observations demonstrate that multi modal signal correlation enables earlier identification of abnormal execution states, including latency drift, resource saturation, and thread contention, when compared to conventional threshold based monitoring systems. The findings indicate measurable improvements in detection lead time and diagnostic precision, reducing the operational gap between failure emergence and corrective action. This study contributes a structured observability architecture and failure anticipation model that advances reliability engineering practices for Java microservices. From an academic perspective, the work extends observability research beyond descriptive monitoring toward predictive operational intelligence. From an industry standpoint, the framework offers a scalable and vendor neutral foundation for building resilient cloud native systems that prioritize service continuity and operational foresight.
DOI: http://doi.org/10.5281/zenodo.17938824
