Back to search

Harokopio University

Models of Computational Infrastructures and Optimal Resource Orchestration

Abstract

dc:description

Containerization has emerged as the predominant paradigm for deploying microservices in modern distributed computing environments, transforming the way applications are packaged, distributed, and managed. By encapsulating applications and their dependencies into lightweight and portable containers, developers can ensure consistent performance across a wide range of platforms and infrastructures. Among the various technologies that support this paradigm, Docker and Kubernetes have become the most widely adopted platforms, forming the foundation of contemporary DevOps practices and cloud-native ecosystems. Despite their popularity, each platform introduces its own domain-specific language, configuration syntax, and conceptual model for defining deployment plans. These differences often create challenges for developers, system architects, and researchers who aim to design, deploy, and manage containerized applications in a consistent and interoperable manner across diverse environments.To address these challenges, this thesis introduces the Containerized Application Deployment Ontology (CADO). CADO is an ontology developed using the Web Ontology Language (OWL) that provides a formal representation of the key concepts, relationships, and terminologies used in containerized deployment models. It establishes a shared semantic foundation that unifies the deployment vocabularies of Docker and Kubernetes, enabling a higher level of interoperability and reducing ambiguity in how deployment structures are defined and interpreted. Through this semantic alignment, CADO facilitates automation, reasoning, and knowledge sharing within the broader containerization ecosystem.Building on this foundation, the thesis also presents the Cloud Edge Application Modeling Language (CEAML), which is a technology-agnostic modeling language inspired by the Topology and Orchestration Specification for Cloud Applications (TOSCA) open standard. CEAML provides a unified framework for describing, orchestrating, and managing microservices across the cloud–edge continuum. It is designed to support the growing demand for distributed intelligence, scalability, and resource optimization at the edge of the network as well as within the cloud. The syntax and structure of CEAML are informed by both state-of-the-art cloud modeling languages and the formal definitions established through the CADO ontology. Together, these two contributions offer a comprehensive solution for bridging the conceptual gap between heterogeneous container technologies and for advancing the modeling of microservice deployments in hybrid computing environments.To demonstrate practical applicability, an interpreter is implemented to translate CEAML specifications into Kubernetes deployment files, serving as a proof of concept for technology-agnostic orchestration. This implementation further illustrates how future interpreters could be extended to support additional container DSLs, functioning within a strategy pattern–based DevOps tool. Finally, CEAML is validated through two real-world platforms, complemented by a monitoring mechanism capable of observing physical machines, virtual machines, and containers. This mechanism supports orchestrators in making informed deployment and management decisions, thereby confirming the practical effectiveness of CEAML in cloud–edge environments.

Degree

thesis:*
Grantor dc:publisher
Harokopio University
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Korontanis, Ioannis
  • Κοροντάνης, Ιωάννης

Subjects

dc:subject × 16

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/62363

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Korontanis, Ioannis; Κοροντάνης, Ιωάννης. Models of Computational Infrastructures and Optimal Resource Orchestration. Harokopio University, 2026. http://hdl.handle.net/10442/hedi/62363