Author: 

Main Authors: Billy Pinheiro, Antonio Abelem (Federal University of Para), Rafael Pasquini (Federal University of Uberlandia), David Moura, Christian Rothenberg (University of Campinas), Lefteris Mamatas, Panagiotis Papadimitriou (University of Macedonia)

Adittional Authors: Sand Luz Correa, Kleber Vieira Cardoso (Federal University of Goias), Fábio Verdi (Federal University of Sao Carlos), Augusto Venâncio Neto (Federal University of Rio Grande do Norte), Alex Galis (University College London), Luis Contreras (Telefonica Investigacion y Desarrollo), Joan Serrat (Telefonica Investigacion y Desarrollo)

Focus Area: 

Cloud Computing

Who stands to benefit and how: 

NECOS enables Service Providers (SP) and Network Operators (NO) to provisioning innovative network functionalities, new services, which need to benefit from the increasing dynamicity of the ICT markets. Thus, the LSDC concept is used to enable Cloud Slicing through the Slice as a Service concept as a unified management framework for both networking and computational domains, providing intelligent orchestration for federated cloud infrastructures.

Position Paper: 

I. INTRODUCTION

In the past decade, the Information and Communication Technology (ICT) sector has experienced rapid changes in both platform scale and application scope. The diversification of service models offered in Cloud Computing (CC), such as: (i) Software as a Service (SaaS), (ii) Platform as a Service (PaaS), and (iii) Infrastructure as a Service (IaaS). This variety of cloud services creates a new challenge for service providers that are using separately managed computing, connectivity, and storage resources to easily deploy new services, as well as enforcing reasonable Service Level Agreements (SLAs).

To provide a single abstraction with end-to-end (E2E) cloud capabilities (processing and storage) and networks, the NECOS platform uses the cloud network slicing concept. This concept has been defined in several ways in the literature, but is generally understood as a physical and virtual resource composition (computing, network and storage) created on demand, having a common underlying infrastructure with independent control, management, and orchestration. These slices must be mutually isolated and flexible enough to accommodate different tenant needs, providing a Slice as a Service model with on-demand virtual infrastructure management (VIM) [1].

II. NECOS

The potential of Cloud Slicing through the Lightweight Slice Defined Cloud (LSDC) is massive, as it focuses on creating a Slice as a Service mechanism which spans multiple cloud infrastructures, from the data center to the edge. One of the main NECOS objective is to formalize the requirements and to provide the integrated management and orchestration mechanisms that need to be used to achieve this goal. As service providers offer more applications on the cloud and as consumer’s demand increases at a non-linear rate, eventually even the largest cloud infrastructure will run out of resources to supply. While clouds can be over-provisioned to handle spikes in demand, this is hugely expensive and not scalable.

The LSDC represents a novel approach for automating the process of optimal cloud configuration by creating the Cloud Slice concept across all resources in a set of federated data centers, as well as providing uniform management of currently separated computing, connectivity, and storage resources [3].

A. NECOS Architecture

The complete NECOS platform architecture presented in Figure 1, defines the components for E2E cloud network slices provisioning in a multi-provider federation. In NECOS, a cloud network slice E2E can be composed of multiple cloud slices provisioned in different federation data centers, and are interconnected by network slices [2].

The Slice Orchestrator performs slice manipulation, that is, it asks the Marketplace for the different parts that will be included in an E2E slice. Additionally, it supervises the service's life cycle management executed in the slices, based on the resources' information collected by the Infrastructure & Monitoring Abstraction (IMA) component.

The IMA component provides an abstract northbound interface that allows the Slice Orchestrator to perform its functions while slice part information remains transparent. To achieve these goals, multiple adapters translate requests from the northbound interface to VIM, abstracting the slice parts heterogeneity at the southbound interface of the IMA.

The discovery of features that will make up an E2E slice is the responsibility of the Marketplace component, which is a distributed system responsible for finding parts of appropriate slices from a set of federated cloud infrastructures. As shown in Figure 1, the Marketplace consists of two components, named Slice Broker and Slice Agent. Slice Broker discovers the slice parts through interaction with a set of Slice Agents hosted in the resource domains involved.

III. CONCLUSIONS AND FUTURE WORK

In this paper, we present the NECOS proposal towards addressing the limitations of current cloud infrastructures by proposing the LSDC approach. LSDC acts as an enabler of Cloud Slicing through the Slice as a Service concept as a unified management framework for both networking and computational domains, providing intelligent orchestration for federated cloud infrastructures.

ACKNOWLEDGEMENTS

This research was partially supported by the H2020 4th EU-BR Collaborative Call, under the grant agreement no. 777067 (NECOS - Novel Enablers for Cloud Slicing), funded by the European Commission and the Brazilian Ministry of Science, Technology, Innovation, and Communication (MCTIC) through RNP and CTIC.

Year: 
2019