| _version_ | 1866901728961495040 |
|---|---|
| author | Manuja Bandal |
| author_facet | Manuja Bandal |
| contents | <p><span>Testing software in cloud environments, especially within AWS infrastructures, presents distinct obstacles due to dynamic resource allocation, decentralized architectures, and unpredictable execution conditions. Conventional testing methods often fail to detect cloud-specific faults such as transient errors, race conditions in autoscaling, and inconsistencies in distributed systems. This paper introduces AI-Enabled Adaptive Fault Injection (AIAFI), a novel autonomous testing framework specifically designed for AWS-based applications. AIAFI autonomously detects, injects, and modifies fault scenarios in cloud-native applications through reinforcement learning (RL) and evolutionary search mechanisms. By utilizing AWS-integrated observability tools such as CloudWatch, X-Ray, and AWS Fault Injection Simulator (FIS), our approach enhances fault detection by 65% compared to leading-edge testing methodologies. Experimental results demonstrate that AIAFI effectively optimizes test execution, minimizes downtime, and improves the resilience of AWS-powered infrastructures.</span></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15044761 |
| institution | Zenodo |
| language | |
| publishDate | 2022 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | AI-Enabled Adaptive Fault Injection for Self-Regulating Software Testing in AWS Cloud Platforms Manuja Bandal Fault Injection Cloud AWS Artificial Intelligence Software Testing <p><span>Testing software in cloud environments, especially within AWS infrastructures, presents distinct obstacles due to dynamic resource allocation, decentralized architectures, and unpredictable execution conditions. Conventional testing methods often fail to detect cloud-specific faults such as transient errors, race conditions in autoscaling, and inconsistencies in distributed systems. This paper introduces AI-Enabled Adaptive Fault Injection (AIAFI), a novel autonomous testing framework specifically designed for AWS-based applications. AIAFI autonomously detects, injects, and modifies fault scenarios in cloud-native applications through reinforcement learning (RL) and evolutionary search mechanisms. By utilizing AWS-integrated observability tools such as CloudWatch, X-Ray, and AWS Fault Injection Simulator (FIS), our approach enhances fault detection by 65% compared to leading-edge testing methodologies. Experimental results demonstrate that AIAFI effectively optimizes test execution, minimizes downtime, and improves the resilience of AWS-powered infrastructures.</span></p> |
| title | AI-Enabled Adaptive Fault Injection for Self-Regulating Software Testing in AWS Cloud Platforms |
| topic | Fault Injection Cloud AWS Artificial Intelligence Software Testing |
| url | https://doi.org/10.5281/zenodo.15044761 |