Explainable Anomaly Detection for HPC Storage during Computational Modelling

Abstract

Scientific-computing storage experiences bursts of parallel reads and checkpoint writes that can resemble genuine infrastructure anomalies. The difficulty is particularly visible when a system is transferred to a new setting. To address this, the study will separate workload-driven congestion from storage faults while maintaining interpretable alerts. The research design centres on a trace-replay experiment using computational-model I/O patterns and injected failures under controlled concurrency. The choice of evaluation units and exclusion criteria is informed by the broader methodological record [3–7]. The analysis will contrast static queue thresholds, robust multivariate detectors and explanation-producing anomaly models. The design prioritises precise operational definitions and a straightforward baseline before introducing more complex alternatives. Related enterprise-storage studies inform the assessment of operational reliability [1] and the deployment assumptions adopted for system-level comparison [2]. Assessment will focus on tail latency, false alerts per workload, fault lead time and root-cause agreement with diagnostic logs. The supporting evidence base motivates the chosen comparator conditions and robustness criteria [8–12]. Critical sensitivity checks address cache warm-up, burst timing, shared-file interference and operator interpretation of alarms. Further systems and modelling literature supplies context for the proposed sensitivity analyses [13–17]. Primary analyses would be reported by the main deployment or participant subgroup so that averaged results do not conceal systematic failures. The expected output is a reusable evaluation framework with transparent uncertainty and reporting rules, informed by wider reproducibility literature [18–23].