QoS-Aware orchestration of applications across the cloud-edge computing continuum is crucial for optimising performance, reducing latency, and sustainably managing resource usage. In this article, we present Avant Intelligent Deployer (AID), the open-source prototype of a novel Kubernetes-based orchestrator for automated declarative application management in Cloud-Edge settings, focusing on environmental sustainability. AID relies on declarative management policies specified in Prolog to determine a deployment node for each application service, considering average carbon intensity, CPU, GPU, and RAM availability, node location and affinity constraints between services. We demonstrate our prototype tool over a small-scale industrial case study.
Declarative application management with Kubernetes
Forti, Stefano
Primo
;Brogi, Antonio
2025-01-01
Abstract
QoS-Aware orchestration of applications across the cloud-edge computing continuum is crucial for optimising performance, reducing latency, and sustainably managing resource usage. In this article, we present Avant Intelligent Deployer (AID), the open-source prototype of a novel Kubernetes-based orchestrator for automated declarative application management in Cloud-Edge settings, focusing on environmental sustainability. AID relies on declarative management policies specified in Prolog to determine a deployment node for each application service, considering average carbon intensity, CPU, GPU, and RAM availability, node location and affinity constraints between services. We demonstrate our prototype tool over a small-scale industrial case study.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


