Abstract
A system predicts imminent network-device failures and initiates causally coherent evidence capture before volatile state disappears. In response to a causal token, device-local agents collect native, capability-aware evidence from selected causal neighbors, fuse relevant change context, and normalize the resulting artifacts into a vendor-agnostic bundle. A tiered reproduction engine uses the bundle to recreate the failure through simulation, virtual-device organic state arrival, or hardware-sandbox execution, and then ablates the reproduction to identify a minimal action set that still produces the failure signature. The system enables reproducible diagnosis of intermittent, multi-device network failures without requiring a central mirror or orchestrator.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
R Devisetty, Kamakshi and Ramu, Nikhil Alampalli, "ARTIFICIAL INTELLIGENCE (AI)-DRIVEN REPRODUCTION OF NETWORK DEVICE FAILURES FROM PRE-FAILURE EVIDENCE CAPTURE", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/11265