Multi-cloud resource management: cloud service interfacing

Abstract

Cloud service abstractions are currently used to hide the underlying complexity given by existing technologies and services, in hope of facilitating the enacting of Cloud Federations and Marketplaces. In particular, resource management systems dealing with multiple Cloud providers need to expose an uniform interface for various services and to build wrappers for the Cloud service APIs. In this paper we discuss the solution adopted by a recent developed open-source and vendor agnostic platform-as-a-service for Multi-Cloud application deployment. The middleware includes a multi-agent system for automatic Cloud resource management. With a modular design, the solution provides a flexible approach to encompass new Cloud service offers as well as new resource types. This paper focuses on the modules which enable resource abstraction and automatized management.

The cloud application modelling and execution language

“…To address the vendor lock-in effects [34], multi-cloud resource management (MCRM) has been proposed [31], which offers organisations several capabilities including [2]: (a) optimal use of best possible cloud services from a variety of offerings supplied by a multitude of cloud providers; (b) ability to sustain an optimal quality level via the application dynamic reconfiguration; (c) ability to achieve a better security level by exploiting suitable security services; (d) ability to move applications near the client location to improve application performance; (e) ability to conform to national and international regulations.…”

Cloud Brokerage

“…Meeting Requirements [Chichin et al 2014] [Dastjerdi et al 2010] [Patiniotakis et al 2014] [Quarati et al 2016] [Yangui et al 2014] [Amato et al 2013b] [Anastasi et al 2017] [den Bossche et al 2013] [Javed et al 2016] [Michon et al 2017] [Pawluk et al 2012] [Weerasiri et al 2015] [Munteanu et al 2014] [Elkhatib et al 2013] [Aleksiev et al 2013] [Ferrer et al 2012] [Yang et al 2012] [Apduhan et al 2015] [Calheiros et al 2012] [Itani et al 2012] [Javadi et al 2012] [Jrad et al 2013] [Kertesz et al 2014] [Kim et al 2015] [Lucas-Simarro et al 2011] [Nair et al 2010] [Sharma et al 2015] [Desair et al 2013] [Paraiso et al 2014 [A a z a m a n d H u h 2 0 1 7 In the following, each category of motivation is described in detail. Dimensionality.…”

“…Pathological Motivations. These motivations vary between achieving specific interoperability guarantees (e.g., between different grid and cloud resources [Aleksiev et al 2013;Munteanu et al 2014;Somasundaram et al 2014]), enabling collaboration between users from different backgrounds [Elkhatib et al 2013], hosting compute-intensive applications [Yang et al 2012], and addressing complex legislative issues [Ferrer et al 2012].…”

Ontology-based knowledge recognition in service-oriented virtual research environments

“…We are analyzing in this research only the VRE functionality on top of real use cases, and by the way make it possible to take into account the privacy aspects. the communication and dissemination strategy of the VRE have a key role into the accomplishment of its main technical objectives by: (1) reporting to Universities and Research Institutes, to the general public and to the media; (2) exploiting the VRE outcomes and results in order to help reinforcing the EU industrial base in the domain of e-infrastructures; (3) communicating about SMART-VRE benefits in VRE's so as to ensure the exploitation.…”

“…High Performance Computation (HPC) is set forward. The current e-infrastructure services related to HPC, Grid and Cloud, which have been funded by national or European funding agencies (like FP7 PRACE for HPC, EGI-Inspire for Grid, BonFIRE for Cloud services), are focused on computational intensive services, rather than on data processing [1], [7].…”