Causal Inference for Quantifying Noisy Neighbor Effects in Multi-Tenant Cloud Environments

2026-04-03

AI Summary

This study addresses the performance unpredictability caused by the “Noisy Neighbor” problem in multi-tenant cloud environments, where co-located workloads interfere with one another due to resource contention, yet lack interpretable quantification and causal analysis. The authors present the first explainable quantification of Noisy Neighbor effects on Kubernetes by integrating controlled experiments with a multi-stage causal inference framework that includes Granger causality analysis. Their approach identifies distinct “degradation signatures” associated with different types of resource contention, effectively transforming the problem into a diagnosable phenomenon. Experimental results demonstrate that I/O-intensive workloads suffer performance degradation of up to 67%, and that causal links increase significantly—by 75%—when noisy neighbors are active. These findings provide critical insights for service-level agreement (SLA) enforcement and intelligent workload scheduling.

Abstract

Resource sharing in multi-tenant cloud environments enables cost efficiency but introduces the Noisy Neighbor problem, i.e., co-located workloads that unpredictably degrade each other's performance. Despite extensive research on detecting such effects, there are no explainable methodologies for quantifying the severity of impact and establishing causal relationships among tenants. We propose an analytical that combines controlled experimentation with multi-stage causal inference and validates it across 10 independent rounds in a Kubernetes testbed. Our methodology not only quantifies severe performance degradations (e.g., up to 67\% in I/O-bound workloads under combined stress) but also statistically establishes causality through Granger causality analysis, revealing a 75\% increase in causal links when the noisy neighbor activates. Furthermore, we identify unique "degradation signatures" for each resource contention vector (i.e., CPU, memory, disk, network), enabling diagnostic capabilities that go beyond anomaly detection. This work transforms the Noisy Neighbor from an elusive problem into a quantifiable, diagnosable phenomenon, providing cloud operators with actionable insights for SLA management and smart resource allocation.

Problem

• Noisy Neighbor
• Causal Inference
• Multi-Tenant Cloud
• Performance Degradation
• Resource Contention

Innovation

• causal inference
• noisy neighbor
• multi-tenant cloud
• resource contention
• performance degradation