kubernetes orchestration

This work proposes CODECO, a framework designed to address the challenges of traditional centralized Kubernetes in federated edge environments, where heterogeneous infrastructure, device mobility, and multi-provider collaboration are prevalent. CODECO enables edge autonomy while preserving global consistency through co-orchestration of data, computation, and networking. It integrates a semantic application model, a partitioned federation mechanism, AI-driven scheduling decisions, and a hybrid governance model. Built upon an extended Kubernetes architecture, CODECO supports context-aware microservice deployment and adaptive management. The framework’s efficacy in orchestrating applications across complex federated edge-cloud scenarios is validated through a reproducible experimental platform, demonstrating its capability to efficiently manage dynamic and heterogeneous edge environments.

To address inefficient task scheduling and resource fragmentation caused by resource constraints and unstable connectivity in edge computing, this paper proposes a QoS-aware cloud-edge collaborative resource management framework. The framework introduces a novel batch-processing scheduling algorithm and a three-tier dynamic migration mechanism—comprising intra-edge, edge-to-cloud, and cloud-to-edge migrations—integrated with lightweight container live migration, QoS-aware scheduling policies, and an edge-state-adaptive decision model, implemented atop an extended Kubernetes platform. Experimental results demonstrate an average 23.6% improvement in edge Pod deployment rate, a 41.2% reduction in deployment instability (measured by standard deviation), a 57.3% decrease in resource fragmentation, and 100% QoS compliance. This work is the first to deeply integrate batch scheduling with multi-level dynamic migration, significantly enhancing both service quality for latency-sensitive applications and resource utilization efficiency in edge environments.

This study addresses the practical disparities and co-evolution between high-performance computing (HPC) and edge computing architectures within the cloud continuum. It presents the first large-scale empirical analysis based on 396 real-world, production-grade AWS architectures. Methodologically, we propose a multidimensional, data-driven framework encompassing service topology identification, storage type classification, architectural complexity quantification, and ML service integration statistics. Results reveal systematic differences—and complementary patterns—between HPC and edge architectures across four dimensions: core service composition (e.g., EC2 versus Greengrass/Lambda), storage design paradigms (parallel file systems versus distributed lightweight caches), complexity distributions, and ML embedding strategies. This work delivers the first industry-scale architectural benchmark for the cloud continuum, providing empirically grounded insights and methodological foundations for cross-domain architecture design, resource optimization, and cloud-native convergence of HPC and edge computing.

Traditional high-performance computing (HPC) and cloud computing have long remained siloed due to divergent origins, cultures, and technological trajectories, hindering their joint ability to address emerging heterogeneous scientific workloads demanding both agile service orchestration and ultra-low-latency, state-aware performance. Method: This paper introduces “Converged Computing,” a novel paradigm enabling co-deployment of the HPC workload manager Flux and user-space Kubernetes (Usernetes) on native supercomputing clusters. Leveraging Linux namespaces, cgroups, and a custom network plugin, it establishes an infrastructure-level convergence architecture. Contribution/Results: The approach unifies cloud-native automation and portability with HPC’s low-latency interconnects, high-bandwidth networking, and fine-grained resource scheduling. Experimental evaluation in hybrid environments demonstrates low-overhead execution of HPC applications and efficient cross-environment communication. An open-source, reproducible deployment framework is provided, offering a practical pathway for HPC centers to adopt cloud-native technologies.

This study addresses prolonged task completion times, low resource utilization, and high resource release latency in Docker/Kubernetes containers on cloud-native platforms running compute-intensive workloads (e.g., big data and deep learning). We systematically evaluate the performance impact of diverse resource scheduling strategies through system-level monitoring—leveraging cgroups and metrics-server—and multi-workload stress testing. For the first time, we empirically quantify how key resource configurations significantly affect task completion time (±79.4% variation) and resource release latency (+116.7% degradation). Based on these findings, we propose an evidence-driven configuration optimization paradigm that reduces maximum task completion time by up to 79.4% and precisely identifies configuration bottlenecks responsible for latency. Our results provide reproducible, transferable empirical foundations for resource management tuning and deployment decisions in cloud-native environments.

To address high energy consumption of containerized applications and the lack of fine-grained energy awareness in resource scheduling within heterogeneous edge-cloud environments, this paper proposes an embedded energy-aware scheduling framework. The framework integrates real-time power consumption metrics across both computation and networking dimensions into the Kubernetes scheduler and implements dynamic energy-efficiency optimization on an ARM-based physical edge testbed. Its key innovations include a lightweight hardware-coordinated monitoring mechanism and a redesigned scheduling decision logic that jointly optimizes workload distribution and energy consumption. Experimental results demonstrate that, under high-load conditions, the proposed approach reduces total system energy consumption by 23.7% compared to vanilla Kubernetes, while maintaining QoS guarantees and high resource utilization—thereby significantly enhancing the energy efficiency of edge-cloud collaborative systems.

This study addresses the challenge of automating the deployment of multi-service containerized applications in heterogeneous edge-cloud environments, where minimizing manual intervention while ensuring performance guarantees remains difficult. The authors evaluate and validate the open-source CODECO toolkit’s orchestration capabilities across diverse hardware platforms—including ARM, AMD, and Raspberry Pi—and lightweight Kubernetes distributions such as k3s. Experimental results demonstrate that, compared to standard Kubernetes workflows, CODECO significantly reduces human intervention during deployment while maintaining competitive deployment efficiency, runtime performance, and manageable resource overhead. These findings highlight CODECO’s strong compatibility and its potential to lower operational complexity in edge-cloud collaborative deployments.

Existing Kubernetes schedulers struggle to simultaneously optimize user-defined QoS objectives—such as energy efficiency, cost, and global performance—while lacking automated, declarative orchestration capabilities across heterogeneous cloud-fog-edge clusters. To address this, we propose the first QoS-aware federated orchestration system. Our approach employs a lightweight, Raft-replicated resource agent architecture tightly coupled with a centralized knowledge repository, enabling, for the first time, automatic translation of user-specified YAML-declared multi-dimensional QoS constraints (e.g., latency, energy consumption, cost) into microservice placement and dynamic migration policies. The system integrates Istio service mesh and federated cluster management to support policy-driven scheduling, QoS-compliant rescheduling, and zero-touch failover. Evaluated on a nine-cluster testbed, our system demonstrates both effectiveness and scalability in meeting diverse, cross-layer QoS requirements.

Data centers’ high energy consumption exacerbates global carbon emissions, while existing Kubernetes carbon-aware scheduling relies on centralized machine learning models—suffering from privacy leakage and poor cross-domain generalizability. To address these challenges, this paper proposes the first federated learning-based energy consumption prediction framework for container orchestration. We extend the Kepler monitoring system using Flower and XGBoost, and introduce FedXgbBagging—a novel federated aggregation strategy tailored for distributed energy-efficiency modeling. This enables privacy-preserving, decentralized collaborative training across organizational boundaries. Evaluated on the SPECPower dataset, our framework reduces mean absolute error by 11.7% compared to centralized baselines, significantly improving prediction accuracy and enhancing the practical feasibility of green, carbon-aware scheduling in production Kubernetes environments.

This study addresses the challenges faced by edge-cloud-native applications in cross-industry adoption, including fragmented toolchains, steep learning curves, and inconsistent performance across hybrid environments. Through in-depth interviews with practitioners from multiple sectors, the work offers the first systematic insight into the real-world pain points experienced by non-technical teams during digital transformation. It reveals that such teams prioritize productivity, service quality, and usability over cost alone. Grounded in qualitative analysis, the research identifies key platform design requirements centered on developer-friendliness, end-to-end lifecycle simplification, and SLA-aware orchestration, with a focus on distributed network computing, hybrid cloud management, and service-level agreement (SLA) assurance. These findings provide a practice-oriented roadmap for the evolution of converged cloud-network infrastructures.