Deployment of TOSCA cloud services archives using Kubernetes

Abstract

In recent years container virtualization and container management emerged in the context of Cloud computing as a new paradigm in IT enterprises. It introduces new approaches that enable the IT industry to manage their application and services more effectively in the Cloud. With the rapid increase of usage of Cloud computing, IT companies introduce new tools to manage their applications in Cloud environments. However, each tool has its own kind of definitions and specifications on describing the applications in their platforms which creates vendor lock-in for its users and also hampers the portability features of Cloud applications. To solve this issue, TOSCA has been introduced to the industry by OASIS. The Topology and Orchestration Specification for Cloud Applications (TOSCA) provides a standardization approach enabling portability of Cloud services between different Cloud Computing providers. The main goal of TOSCA is to model enterprise applications in a standardized and technology-independent way regardless of a specific Cloud provider or environment. To model and deploy applications using TOSCA properly, all required artifacts are packaged and bundled as TOSCA Cloud Service Archives (CSARs). Such CSARs are then used by Cloud orchestration engines to deploy the application to Cloud platforms. At the technology level, several deployment and cluster management approaches and tools are rapidly emerging such as Docker Compose, Docker Swarm, Kubernetes, Nomad, and Apache Mesos. Most of them are centered around containerization of middleware and application components. The focus of this thesis is to provide mapping concepts of TOSCA application topologies to a container-based deployment and management approach. Since Kubernetes and Docker are the most prominent open-source solutions in this field, we specifically consider Kubernetes and Docker as part of the prototype implementation. To assess feasibility of the proposed approach and usability of the system, we also provide case studies based on a motivating scenario.