Barletta, Marco (2024) Mixed-criticality Orchestration of Real-time Containerized Systems. [Tesi di dottorato]
![[thumbnail of Barletta_PhDDissertation_signed.pdf]](https://www.fedoa.unina.it/style/images/fileicons/other.png "Barletta_PhDDissertation_signed.pdf") |
Documento PDF
Barletta_PhDDissertation_signed.pdf Visibile a [TBR] Amministratori dell'archivio Download (15MB) | Richiedi una copia |
| Tipologia del documento: | Tesi di dottorato |
|---|---|
| Lingua: | English |
| Titolo: | Mixed-criticality Orchestration of Real-time Containerized Systems |
| Autori: | Autore Email Barletta, Marco marco.barletta@unina.it |
| Data: | 11 Dicembre 2024 |
| Numero di pagine: | 208 |
| Istituzione: | Università degli Studi di Napoli Federico II |
| Dipartimento: | Ingegneria Elettrica e delle Tecnologie dell'Informazione |
| Dottorato: | Information technology and electrical engineering |
| Ciclo di dottorato: | 37 |
| Coordinatore del Corso di dottorato: | nome email Russo, Stefano stefano.russo@unina.it |
| Tutor: | nome email Cinque, Marcello [non definito] |
| Data: | 11 Dicembre 2024 |
| Numero di pagine: | 208 |
| Parole chiave: | containers, orchestration, real-time, mixed-criticality, resilience, cloud |
| Settori scientifico-disciplinari del MIUR: | Area 09 - Ingegneria industriale e dell'informazione > ING-INF/05 - Sistemi di elaborazione delle informazioni |
| Informazioni aggiuntive: | Appartengo al ciclo XXXVII, che non compariva tra le opzioni |
| Depositato il: | 28 Dic 2024 10:03 |
| Ultima modifica: | 12 Ago 2026 05:38 |
| URI: | https://www.fedoa.unina.it/id/eprint/16490 |
Abstract
As critical computing systems employed in industry verticals grow increasingly complex, hardware components are replaced with software ones that demand greater computational resources. Hence, embedded systems are evolving into more powerful mixed-criticality platforms. Recent research has explored edge computing, virtualization, and cloud-native technologies to automate component management and increase flexibility, interoperability, and scalability. However, cloud-native technologies like container orchestration systems were not designed to meet the heterogeneity of hardware and non-functional requirements of industrial settings. This dissertation investigates whether current container orchestration systems meet the requirements of critical systems, and introduces mixed-criticality orchestration to address their limitations. The dissertation makes three key contributions: i) it introduces a model for mixed-criticality orchestration, ii) it performs a failure and timing analysis to assess the behavior of container orchestration systems in non-nominal conditions, and iii) it proposes a set of solutions based on the model to improve the resilience, timeliness, and isolation from interference for both container orchestration and containers. The analysis reveals that even a single error can cause overloads and disrupt an entire cluster, whereas high orchestration loads can cause delays of tens of seconds in scaling services or handling failures, threatening service level objectives (SLOs). The proposed solutions include designs and methods for mixed-criticality orchestration, which builds upon the concepts of node assurance and service and pod criticality to differentiate the management of critical services when orchestrating, placing, and running them. The implemented prototypes demonstrate that mixed-criticality orchestration improves the resilience of critical services by providing stable orchestration times, and improved timing and failure isolation for critical containers.
Downloads
Downloads per month over past year
Actions (login required)
 |
Modifica documento |
