Alibaba Microservices Trace v2021 Stress-Warning Pipeline: Reproducibility Package
Year: 2026
Type: software
Abstract: Reproducibility materials for the study “When Topology Is Not Enough: Confirmatory Evaluation of Local and Graph-Based Warnings for Microservice Response-Time Stress.” This release provides methods-focused software and documentation supporting a leakage-controlled, temporally separated evaluation of local and topology-aware models for five-minute-ahead microservice response-time stress warning using Alibaba Microservices Trace v2021. The package documents the analytical workflow, source-data acquisition and provenance, frozen research design, temporal development and locked-evaluation structure, model and warning-governance configurations, validation utilities, and code supporting the reproducibility of the study. The evaluated framework distinguishes observed logical service-call (CALLS) relations from physical service-to-host placement (RUNS_ON) relations and maintains a strict separation between development, calibration, pre-test freezing, and final locked evaluation. This public release is intentionally methods-focused. It does not redistribute Alibaba Microservices Trace v2021, source-derived row-level data, manuscript files, publication tables or figures, aggregate research-result files, model checkpoints, prediction-level records, or bootstrap-replicate records. Users must obtain the source trace independently from the official Alibaba Cluster Trace Program and comply with the provider’s applicable access and use terms. The MIT License applies only to original code and documentation created by the authors and included in this release. It does not apply to Alibaba Microservices Trace v2021 or to third-party software dependencies, which remain subject to their respective terms and licenses. Version: 0.1.0Release date: 4 September 2026Authors: Charuay Savithi, Lersak Phothong, and Anupong SukprasertAffiliation: Mahasarakham Business School, Mahasarakham University, Maha Sarakham, ThailandFunding: Financially supported by Mahasarakham University; no grant number is applicable.
Institution: Mahasarakham University
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7208815910