Other Paper INMISC-2015-05

BibliographyWagner, Frank; Mega, Cataldo; Waizenegger, Tim; Raskin, Vadim; Kukhtichev, Sergey: A Practical Approach to Model-Based Cloud Service Deployment: Using a Smart Interpreter and a Domain Specific Language.
In: Proceedings of the 3rd International IBM Cloud Academy Conference ICACON 2015.
University of Stuttgart, Faculty of Computer Science, Electrical Engineering, and Information Technology.
pp. 1-10, english.
IBM, May 2015.
Other Paper.
Corporation3rd International IBM Cloud Academy Conference ICACON 2015
CR-SchemaH.4.0 (Information Systems Applications General)
KeywordsCloud Computing; Service Modeling; Deployment Automation; Domain Specific Language
Abstract

Cloud Service deployment automation is a crucial component of the Cloud Computing concept, because it enables us to create new instances of a service with minimal cost and effort. Cloud service instances consist of many platform and infrastructure components as well as application components, which all need to be deployed and configured. Manually deploying new instances hinders development, it increases the cost of onboarding new customers, and operating large installations becomes impractical and error prone. Deployment automation has long been a practice among software engineers, but solutions like installers, scripts, Chef or Puppet are not suited for the diverse Cloud environments. For this reason, higher-level model-based approaches, like Amazon CloudFormation, Openstack Heat and TOSCA, have emerged. They allow automating a wide variety of services and components, and encourage reusing modelling artifacts. However, these modelling languages are complex and lack tool-support. In practice, their reusability is also limited, because the models are often tightly coupled and monolithic. For these reasons, the model-based deployment approaches have not found wide spread adoption yet, especially in the enterprise segment. Therefore, we propose a different approach to model-based deployment automation: We focus on a specific application domain and define a domain specific language (DSL) for this domain. Then we provide a smart interpreter that translates the high level DSL into an executable service deployment model.

Full text and
other links
Presentation slides
Contacttim.waizenegger@ipvs.uni-stuttgart.de
Department(s)University of Stuttgart, Institute of Parallel and Distributed Systems, Applications of Parallel and Distributed Systems
Project(s)TPL
Entry dateNovember 30, 2015
   Publ. Department   Publ. Institute   Publ. Computer Science