Causal Inference for Root-Cause Diagnosis in Global Infrastructure:  Distinguishing Causes from Correlated Metrics Under Timing Uncertainty

Saketh Gowerneni, Pradeep Kandepaneni, Bhanuprakash Suravarapu, Vasu Babu Narra, Satya Sampreeth Boppudi

Causal Inference for Root-Cause Diagnosis in Global Infrastructure: Distinguishing Causes from Correlated Metrics Under Timing Uncertainty

Authors

Saketh Gowerneni, Pradeep Kandepaneni, Bhanuprakash Suravarapu, Vasu Babu Narra, Satya Sampreeth Boppudi

Keywords:

causal inference, root-cause analysis, observability, distributed systems, do-calculus, structural causal models, clock skew, counterfactual reasoning

Abstract

Modern observability platforms identify whichinfrastructure metrics move together but cannot establish
whether one metric causes another, leaving operators todiagnose failures by manual inspection of correlated signals— a costly delay

References

Information Technology Intelligence Consulting, "ITIC 2024 hourly cost of downtime report," ITIC, 2024. [Online]. Available: https://itic-corp.com/itic-2024hourly-cost-of-downtime-report/

Downloads

Published

2024-12-24

How to Cite

Saketh Gowerneni, Pradeep Kandepaneni, Bhanuprakash Suravarapu, Vasu Babu Narra, Satya Sampreeth Boppudi. (2024). Causal Inference for Root-Cause Diagnosis in Global Infrastructure: Distinguishing Causes from Correlated Metrics Under Timing Uncertainty . Journal of Computational Analysis and Applications (JoCAAA), 33(08), 9107–9112. Retrieved from https://www.eudoxuspress.com/index.php/pub/article/view/5639