ARC-V: Vertical Resource Adaptivity for HPC Workloads in Containerized Environments

2025-05-05

AI Summary

To address the mismatch between the dynamic memory demands of HPC workloads and cloud-native vertical scaling mechanisms in containerized environments, this paper proposes an adaptive scheduling framework for node-level memory elasticity. The method integrates HPC memory consumption pattern modeling with Kubernetes’ runtime resource hot-update capability—enabling, for the first time, eviction-free, fine-grained memory scaling. Built upon the Vertical Pod Autoscaler (VPA) extension architecture, the framework incorporates a memory behavior analysis model, real-time per-Pod memory hot-updates, and node-level resource coordination. Experimental evaluation demonstrates that, compared to standard VPA, the framework reduces average memory usage by 37%, eliminates out-of-memory (OOM) failures entirely, and robustly supports elastic execution of nine representative HPC applications.

Technology Category

• Machine Learning: Hardware-aware ML
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Virtualization and resource management in Web systems and infrastructures
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Existing state-of-the-art vertical autoscalers for containerized environments are traditionally built for cloud applications, which might behave differently than HPC workloads with their dynamic resource consumption. In these environments, autoscalers may create an inefficient resource allocation. This work analyzes nine representative HPC applications with different memory consumption patterns. Our results identify the limitations and inefficiencies of the Kubernetes Vertical Pod Autoscaler (VPA) for enabling memory elastic execution of HPC applications. We propose, implement, and evaluate ARC-V. This policy leverages both in-flight resource updates of pods in Kubernetes and the knowledge of memory consumption patterns of HPC applications for achieving elastic memory resource provisioning at the node level. Our results show that ARC-V can effectively save memory while eliminating out-of-memory errors compared to the standard Kubernetes VPA.

Problem

• Analyzing HPC workload memory patterns in containers
• Identifying inefficiencies in Kubernetes VPA for HPC
• Proposing ARC-V for elastic memory provisioning

Innovation

• Leverages in-flight resource updates in Kubernetes
• Uses HPC memory consumption pattern knowledge
• Enables elastic memory provisioning at node level