# Localizing Faults in Cloud Systems

arXiv: https://arxiv.org/abs/1803.00356 · HTML: https://mlsanity.com/paper/1803.00356/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Mar 2018.

## Built on by (most cited first, 8 of 8)

- [AI for IT Operations (AIOps) on Cloud Platforms: Reviews, Opportunities and Challenges](https://mlsanity.com/paper/2304.04661/) (2023)
- [The PetShop Dataset -- Finding Causes of Performance Issues across Microservices](https://mlsanity.com/paper/2311.04806/) (2023)
- [A Comprehensive Survey on Root Cause Analysis in (Micro) Services: Methodologies, Challenges, and Trends](https://mlsanity.com/paper/2408.00803/) (2024)
- [A Causal Approach to Detecting Multivariate Time-series Anomalies and Root Causes](https://mlsanity.com/paper/2206.15033/) (2022)
- [Learning Dependencies in Distributed Cloud Applications to Identify and Localize Anomalies](https://mlsanity.com/paper/2103.05245/) (2021)
- [MetaCaDI: A Meta-Learning Framework for Causal Discovery from Multiple Environments with Unknown Interventions](https://mlsanity.com/paper/2510.22298/) (2025)
- [Unsupervised KPIs-Based Clustering of Jobs in HPC Data Centers](https://mlsanity.com/paper/2312.06546/) (2023)
- [FC-ADL: Efficient Microservice Anomaly Detection and Localisation Through Functional Connectivity](https://mlsanity.com/paper/2512.00844/) (2025)

## Abstract (arXiv)

By leveraging large clusters of commodity hardware, the Cloud offers great opportunities to optimize the operative costs of software systems, but impacts significantly on the reliability of software applications. The lack of control of applications over Cloud execution environments largely limits the applicability of state-of-the-art approaches that address reliability issues by relying on heavyweight training with injected faults. In this paper, we propose emph(LOUD, a lightweight fault localization approach that relies on positive training only, and can thus operate within the constraints of Cloud systems. LOUD relies on machine learning and graph theory. It trains machine learning models with correct executions only, and compensates the inaccuracy that derives from training with positive samples, by elaborating the outcome of machine learning techniques with graph theory algorithms. The experimental results reported in this paper confirm that LOUD can localize faults with high precision, by relying only on a lightweight positive training.