1,720,969 research outputs found
Soluzioni di Monitoraggio per Sistemi Distribuiti su Larga Scala
No full textL'interesse crescente per la gestione di sistemi Ultra-Large-Scale è stato il principale fattore trainante per la ricerca e lo sviluppo di nuove soluzioni e architetture di monitoraggio.
I moderni data center, che spesso supportano servizi basati su cloud, sono caratterizzati da un numero enorme di risorse hardware e software che collaborano spesso in modi complessi e imprevedibili. Comprendere lo stato di questi sistemi al fine di supportare l'analisi prestazionale, la gestione del carico di lavoro, la capacità di produzione e il rilevamento dei guasti, richiede soluzioni di monitoraggio scalabili che dovrebbero raccogliere e valutare continuamente grandi flussi di dati in tempo quasi reale.
Le soluzioni di monitoraggio odierne, sia proprietarie che open-source, presentano spesso alcuni aspetti negativi, tra cui carenza di scalabilità, scarsa rappresentatività delle condizioni di stato globali del sistema, incapacità nel garantire la persistenza nella fornitura dei servizi, e limitazioni nel monitoraggio di applicazioni multi-tenant. Inoltre, queste soluzioni di monitoraggio spesso non riescono a gestire grandi flussi di dati, ovvero a monitorare ad alte frequenze di campionamento le quali causano elevati costi computazionali e di comunicazione, per la raccolta, l'archiviazione e la gestione delle informazioni.
Questa tesi valuta e propone soluzioni di monitoraggio innovative, che spaziano dai sistemi embedded, fino alla gestione di grandi flussi di dati in sistemi Ultra-Large-Scale.
In particolare, forniamo soluzioni efficaci per la raccolta, l'archiviazione e la gestione di grandi flussi di dati proponendo:
(i) architetture scalabili e robuste per monitorare numerose risorse che interagiscono con fonti eterogenee;
(ii) algoritmi di gestione che decidono in modo adattativo la ridistribuzione trasparente delle sessioni live di macchine virtuali in sistemi su larga scala;
(iii) nuove architetture che, combinando un approccio gerarchico con soluzioni decentralizzate, affrontano la sfida del monitoraggio intra-cluster in sistemi su larga scala;
(iv) algoritmi adattativi per il monitoraggio di grandi flussi di dati che migliorano la scalabilità e garantiscono un'elevata affidabilità nel catturare rilevanti variazioni di carico;
(v) nuove architetture di monitoraggio ibride che forniscono elevata scalabilità, efficienza e resilienza, nonché la possibilità monitorare servizi dislocati su diversi cluster e data centers in sistemi Ultra-Large-Scale.The growing interest on resource management of Ultra-Large-Scale systems has engendered novel research activities on adequate monitoring architectures and solutions.
Modern data centers, possibly supporting cloud-based services, are characterized by a huge number of hardware and software resources often cooperating in complex and unpredictable ways. Understanding the state of these systems in order to support performance analysis, workload management, capacity planning and fault detection, requires scalable monitoring solutions that should gather and evaluate continuously large flows in almost real-time.
Existing monitoring solutions are affected by lack of scalability, scarce representativity of global state conditions, inability in guaranteeing persistence in service delivery, and the impossibility of monitoring multi-tenant applications.
Moreover, most architectures fail in managing big data monitoring samples, which cause high computational and communication overheads in collecting, storing, and managing information.
This thesis addresses the challenge by proposing and evaluating innovative monitoring solutions ranging from embedded systems to distributed resource management that are suitable for Ultra-Large-Scale systems. In particular, we provide effective solutions for collecting, storing, and managing big data streams by proposing:
(i) scalable and robust architectures to monitoring multiple resources interacting with heterogeneous sources;
(ii) management algorithms which decide in an adaptive way the transparent reallocation of live sessions of virtual machines in large-scale systems;
(iii) a novel monitoring architecture that, by combining a hierarchical approach with decentralized monitors, addresses the challenge of intra-cluster monitoring in large-scale systems;
(iv) an adaptive algorithm for big data monitoring that improves scalability and guarantees high reliability in capturing relevant load changes;
(v) a novel hybrid monitoring architecture that strives to obtain high scalability, effectiveness and resilience, as well as the possibility of monitoring services spanning across different clusters or even different data centers in Ultra-Large-Scale systems
Fluid Computing & Digital Twins for intelligent interoperability in the IoT ecosystem
No full textThe integration of physical and digital systems is fundamental to enabling intelligent, adaptive, and scalable solutions in modern IoT environments. This paper explores Fluid Digital Twins (FDTs), a novel framework combining Fluid Computing (FC) principles with Digital Twin (DT) technology, to address challenges related to interoperability, dynamic functionality, and adaptability in IoT ecosystems. FC introduces a paradigm shift, enabling seamless data and computational task flow across heterogeneous environments, dynamically adjusting to resource availability and system needs. This paper focuses on embedding intelligence within FDTs to enhance interoperability and enable IoT applications to adapt to changes across both physical and digital domains. By integrating intelligent interoperability mechanisms, FDTs ensure smooth data alignment and compatibility across platforms, adapting to both physical and digital changes. The proposed framework has been implemented, prototyped, and evaluated in the Modena Automotive Smart Area (MASA), a smart city testbed. The evaluation demonstrates FDTs' ability to enhance smart mobility, optimize transportation systems, and provide actionable insights, highlighting their transformative potential in dynamic, data-rich environments. The results emphasize the practical applicability of FDTs in addressing real-world challenges and advancing the capabilities of IoT-driven smart cities
Fluid Computing in the Internet of Things: A Digital Twin Approach
No full textThe concept of Fluid Computing entails a dynamic resource allocation approach, enabling seamless task migration between computing nodes. This paper investigates the fusion of Fluid Computing principles with the Internet of Things (IoT) and introduces the concept of Fluid Digital Twins (FDTs) i.e. cyber-physical entities that bridge the complexities of this integration. FDTs serve as intermediaries, overseeing fluid task migration, optimizing resource use, and simplifying interactions for external digital applications. The paper delves into challenges arising from this fusion, including limited IoT device capabilities, fragmentation, and the necessity of an intelligent intermediary layer. This research article models and presents FDT mechanics, features a prototype with experimental evaluation and concludes by discussing findings and potential future research directions
Towards Operator Digital Twins in Industry 5.0: Design Strategies & Experimental Evaluation
No full textThe concept of Industry 5.0 is set to revolutionize the landscape of modern manufacturing, emphasizing human-centricity and elevating the well-being of industry workers as a central tenet of the production process. This paper extends this vision by integrating the dimension of health, focusing not only on the well-being of the operator but also on the detection of their health condition, predicting potential issues, and consequently enhancing their overall welfare. Building upon this enhanced perspective, our work explores the role of Operator Digital Twins (ODTs), which are instrumental in creating a symbiotic relationship between human operators and industrial machinery. ODTs act as digital counterparts, reflecting the physical and cognitive states of operators, thus facilitating real-time monitoring of their capabilities, workload, stress levels, and various health-related parameters. The paper delves into the motivations driving the development of ODTs, abstractly models their functions, and outlines the architectural blueprint. We present an initial ODT prototype with wearable technology and simulated data together with a discussion of the experimental insights and outcomes
Towards Coordinating Machines and Operators in Industry 5.0 through the Web of Things
No full textThis paper proposes a groundbreaking architecture that reimagines Industry 5.0, emphasizing human-centric technological integration via the Web of Things (WoT) standard. Our approach innovatively digitizes human operators and machinery, creating a responsive industrial ecosystem attentive to real-time human conditions. Central to this is the Operator Thing (OT), a digital replica representing the human operator's status and needs. This system not only recognizes operator stress and discomfort but intelligently adjusts, ensuring optimal human-machine synergy. Our methodology extends to redefining operational parameters and tasks in response to human states, balancing well-being with production efficiency. The ultimate goal is a transformative, adaptive, and empathetic Industry 5.0 environment, validated through rigorous interdisciplinary evaluation
Digital Twins & Fluid Computing in the Edge-to-Cloud Compute Continuum
No full textThis work explores the integration and experimental evaluation of Fluid Computing principles with the Internet of Things (IoT) through the concept of Fluid Digital Twins (FDTs). They have been recently introduced as a cyberphysical paradigm designed to serve as intermediate software components aiming to enable seamless task migration, optimize resource utilization, and streamline interactions. Expanding upon this investigation, the research investigates FDTs within the context of the edge-to-cloud compute continuum. It models and explores the feasibility and ramifications of deploying and orchestrating FDTs and their dynamic capabilities across diverse computational facilities, from edge devices to cloud infrastructure. The paper outlines a new distributed FDT's modeling, presents the implemented prototype within a target reference use case together with its experimental evaluations, and analyzes challenges and opportunities inherent in this dynamic integration
Adaptive, scalable and reliable monitoring of big data on clouds
No full textReal-time monitoring of cloud resources is crucial for a variety of tasks such as performance analysis, workload management, capacity planning and fault detection. Applications producing big data make the monitoring task very difficult at high sampling frequencies because of high computational and communication overheads in collecting, storing, and managing information. We present an adaptive algorithm for monitoring big data applications that adapts the intervals of sampling and frequency of updates to data characteristics and administrator needs. Adaptivity allows us to limit computational and communication costs and to guarantee high reliability in capturing relevant load changes. Experimental evaluations performed on a large testbed show the ability of the proposed adaptive algorithm to reduce resource utilization and communication overhead of big data monitoring without penalizing the quality of data, and demonstrate our improvements to the state of the art.Real-time monitoring of cloud resources is crucial for a variety of tasks such as performance analysis, workload management, capacity planning and fault detection. Applications producing big data make the monitoring task very difficult at high sampling frequencies because of high computational and communication overheads in collecting, storing, and managing information. We present an adaptive algorithm for monitoring big data applications that adapts the intervals of sampling and frequency of updates to data characteristics and administrator needs. Adaptivity allows us to limit computational and communication costs and to guarantee high reliability in capturing relevant load changes. Experimental evaluations performed on a large testbed show the ability of the proposed adaptive algorithm to reduce resource utilization and communication overhead of big data monitoring without penalizing the quality of data, and demonstrate our improvements to the state of the art
Digital Twin Driven Collaboration in Industry 5.0
No full textThis paper explores the integration of Digital Twins (DTs) in Industry 4.0 and 5.0, highlighting their role in enhancing intelligent, collaborative industrial ecosystems. By representing processes, machinery, operators, and products, DTs enable comprehensive life-cycle support and improved shop-floor operations. Intelligent applications and services can harness DTs as structured and interoperable virtual replicas, entrusted with the responsibility of interfacing with the physical world and facilitating access and mediation of interactions therein. Our study proposes structured DT modeling in industrial ecosystems to demonstrate how DTs enable an effective decoupling of responsibilities and capabilities supporting precise monitoring and data synthesis, optimizing production workflows and maintenance. We discuss DTs’ potential in industrial quality control, highlighting efficiency gains and operational improvements in electric motor production through case studies
- …
