<![CDATA[This study investigates whether decomposing a Java-based monolithic application into an event-driven serverless design leads to measurable improvements in resilience and recovery behavior. The evaluation used reliability metrics aligned with predefined service-level objectives, including recovery time objective (RTO), recovery point objective (RPO), mean time to recovery (MTTR), and mean time between failures (MTBF). The results indicate that the serverless variant maintains more stable latency as load increases and recovers from injected failures significantly faster than the monolithic baseline, with MTTR reduced by approximately a factor of two. In addition, execution time and memory limits enforced by the serverless runtime shaped the design by requiring long-running work to be decomposed into smaller stages, which helped confine failures to bounded recovery events and keep error rates within the planned error budget. These findings provide practical guidance for engineers designing reliable serverless systems evaluated against explicit service-level objectives]]>
Copyrights © 2026
