Search

The concept of network slicing enables operators to provision multiple virtual networks on top of a single (shared) physical infrastructure. Adding elasticity to slicing, i.e., the ability to on-demand provision/release dedicated network resources, improves resource utilization. However, efficiently allocating and scaling slice resources, while maintaining state consistency, is challenging. Especially with P4-programmab1e network devices that process packets at Tbps speeds, controller-driven scaling of network functions would be too time-consuming, and data-plane scaling is needed. In this paper, we address this need, by developing a custom scaling protocol and framework that can consistently, with negligible delay, scale network slices and functions transparently to the slice end-users. We compare, via emulation and experiments on programmable hardware, our approach to state-of-the-art scaling techniques and demonstrate significant slice resource utilization improvements and scaling duration reductions.

Networks often need to concurrently process mil- lions of flows with varying Quality-of-Service (QoS) require- ments. Doing so by deploying flow-specific rules at network nodes would require significant memory and overhead.In this paper, we take a fundamentally different approach, called P4QoS, by embedding QoS requirements in the packets themselves and leveraging P4-programmable network switches to process the traffic based on them. We illustrate and evaluate our approach with latency as our QoS metric, but our concept can be applied to other metrics as well. Our evaluation, both in software (Mininet) and in hardware (Intel Tofino), shows that P4QoS can satisfy application-specific QoS requirements with negligible memory overhead.

Network Services (NSs) to be supported over the upcoming 5G infrastructure are envisioned to be very heterogeneous demanding different network functions (NFs) such as Deep Packet Inspection, mobile Evolve Packet Core functions, etc. with various levels of QoS (in terms of bandwidth, latency, etc.). To this end, virtualization techniques as being defined by ETSI Network Function Virtualization (NFV) architecture provides an appealing framework and context where effectively deploying such NSs. Specifically, NFV virtualizes NFs over general-purpose servers hosted in datacenters (DCs). DCs may have different size (in terms of the equipped IT resources) which can be located at different geographical locations being inter-connected via a Wide Area Network (WAN) infrastructure. Those NFs to be deployed in DCs are commonly referred to as Virtual Network Functions (VNFs). In general, specific NSs may require that a number of VNFs deployed potentially in different DCs are sequenced to form the entire targeted service. The selection criteria to choose the DCs where the NS’VNFs are instantiated is diverse, mostly depending on the NS requirements and/or an adopted policy/strategy of the cloud/network operator. For the sake of completeness, such criteria may aim at selecting DCs which minimize the end-to-end latency of the demanded NSs, attain an efficient utilization of the cloud resources (e.g.,load balancing), etc. The ordered set of VNFs and their connectivity constituting NSs is commonly termed as VNF Forwarding Graphs (VNFFG) [1].

Over the last couple of years, industry operators' associations issued requirements towards an end-to-end management and orchestration plane for 5G networks. Consequently, standard organisations started their activities in this domain. This article provides an analysis and an architectural survey of these initiatives and of the main requirements, proposes descriptions for the key concepts of domain, resource and service slicing, end-to-end orchestration and a reference architecture for the end-to-end orchestration plane. Then, a set of currently available or under development domain orchestration frameworks are mapped to this reference architecture. These frameworks, meant to provide coordination and automated management of cloud and networking resources, network functions and services, fulfil multi-domain (i.e. multi-technology and multi-operator) orchestration requirements, thus enabling the realisation of an end-to-end orchestration plane. Finally, based on the analysis of existing single-domain and multi-domain orchestration components and requirements, this paper presents a functional architecture for the end-to-end management and orchestration plane, paving the way to its full realisation. This work was partially supported by the ICT14 5GExchange (5GEx) innovation project (grant agreement no.671636) co-funded by the European Union under the Horizon 2020 EU Framework Programme. Publicado

RESUMO: “Torino Mad Pride” e “Lunatics at Piecework” são duas iniciativas organizadas, seguindo caminhos informais, por um grupo de consumidores/sobreviventes, independentes da Saúde Mental ou outras instituições. Isso os torna únicas no panorama dos movimentos de consumidores locais, ocupados por profissionais e familiares em importantes funções. O objetivo é dar suporte às pessoas afetadas por mal-estar psiquiátrico e organizar iniciativas para combater o estigma. O cerne da ideia de S., que durante uma fase de delírio maníaco viu uma multidão de pessoas reunir-se nas ruas da cidade, seguindo-o. Originalmente, ele apenas queria fazer um filme baseado nessa ideia, para apresentar em festivais de cinema. A partir desse núcleo, desenvolveu-se uma ideia mais complexa e a primeira parada “Torino Mad Pride” foi organizada. Foi em junho de 2011, muita gente envolvida (utentes dos serviços, trabalhadores da área da saúde mental, familiares, organização de utentes, etc.). A participação popular foi inesperada e até o prefeito da cidade participou. Com o tempo, a ideia evoluiu, o projeto “Lunatics at Piecework” (o embrião de uma cooperativa “informal” envolvendo cerca de trinta pessoas em trabalhos temporários) começou a tomar forma, a parada anual tem sido regularmente organizada, e uma revista dedicada a artigos sobre tópicos relacionados com Saúde Mental tem sido irregularmente publicada. Em novembro de 2014 alguns membros da “Lunatics at Piecework” contataram-me, como a Diretora da Unidade de Integração Profissional do Departamento de Saúde Mental do local onde eu trabalhava, para discutir uma parceria, em termos de suporte financeiro por intermédio da concessão de subsídios para dar vida a uma cooperativa social, e para apontar a “Lunatics at Piecework” para realização de trabalhos em favor dos pacientes sob a responsabilidade do Departamento de Saúde Mental (mudança de casa, pequenas remodelações, etc.). Este estudo revisita a história de tais projetos, coligida das vozes dos principais participantes e de importantes espectadores e analisa as razões dos seus altos e baixos e as dificuldades atuais, bem como o que deveria ser feito para dar a estes projetos novas energias. ABSTRACT: “Torino Mad Pride” and “Lunatics at Piecework” are two initiatives put in place following informal paths by a group of consumers/survivors, independent from the Mental Health or other institutions. This renders them unique in the panorama of the local consumers movements that see in prominent roles professionals or family members. They aim to give support to people affected by psychic “malaise” and organize initiatives to fight stigma. They stem from the idea of S., who during a manic delusional phase saw masses of people gathering in the streets of the city following him. Originally, he only wanted to make a movie out of this idea, to present to film festivals. Then a more complex idea took shape, and the first Torino Mad Pride parade was organized. It was June 2011, a lot of people were involved (service users, mental health workers, family members, users’ organizations, etc.). The participation was unexpected, and even the mayor of town paraded. With time the idea has evolved, the “Lunatics at Piecework” project (the embryo of an “informal” cooperative involving about thirty people in temporary jobs) has taken on legs, the annual parade has been regularly organised, a magazine dedicated to writings on topics related to mental health has been irregularly released. In November 2014 some members of LaP contacted me as the director of the Job Placement Unit of the Mental Health Department where I worked, to discuss a collaboration in terms of financial support through the assignment of grants to give life to a social cooperative, and of appointment of LaP to create jobs in favour of patients in charge of the Mental Health Department (house removals, minor refurbishments, etc.) This study revisits the history of these projects, collected from the voices of the key players and of some significant bystanders, and analyses the reasons behind the ups and downs and the present difficulties, and what should be done to give the projects new energies.

The 'good tourism' is capable of generating development in terms of cohesion and sustainability in the territory where it is located. The literature for a long period, has considered the good tourisms in the same way as all other goods. The aim of this study is to demonstrate that by using a different definition of 'good tourism', Cultural Heritage, and appropriate mechanisms for management of the 'Great Cultural Attractions', territorial governance, it is possible to exploit the full potential of the asset. The territorial governance model proposed is bottom-up which includes the participation of all stakeholders in the area will be applied to the Campania Region which has five UNESCO sites. This approach should be able to create a virtuous cycle of growth in the region.

Market fragmentation has resulted in a multitude of network and cloud/data center operators, each focused on different countries, regions and technologies. This makes it difficult and costly to create infrastructure services spanning multiple domains, such as virtual connectivity or compute resources. In this article, we discuss the goals and work being done within the 5GEx (5G Exchange) project in realising a Europe-wide multi-domain platform. This platform aims at enabling cross-domain orchestration of services over multiple administrations or over multi-domain single administrations in the context of emerging 5G networking. The 5GEx vision is based on introducing a unification via network function virtualisation/software-defined networking compatible multi-domain orchestration for networks, clouds and services. We describe the motivation and 5GEx vision, the adopted architecture and the next steps in terms of implementation and experimentation. This work is performed in the framework of the H2020-ICT-2014 project 5GEx (Grant Agreement no. 671636), which is partially funded by the European Commission.

Service meshes factor out code dealing with inter-micro-service communication. The overall resilience of a cloud application is improved if constituent micro-services return stale data, instead of no data at all. This paper proposes and implements application agnostic caching for micro services. While caching is widely employed for serving web service traffic, its usage in inter-micro-service communication is lacking. Micro-services responses are highly dynamic, which requires carefully choosing adaptive time-to-life caching algorithms. Our approach is application agnostic, is cloud native, and supports gRPC. We evaluate our approach and implementation using the micro-service benchmark by Google Cloud called Hipster Shop. Our approach results in caching of about 80% of requests. Results show the feasibility and efficiency of our approach, which encourages implementing caching in service meshes. Additionally, we make the code, experiments, and data publicly available.

none 6 no One of the main challenges of cloud-based service provisioning is to deploy a coordinated control of both application- and network-layer resources in order to provide adaptive service data delivery and adequate user service experiences. In this work we propose a signaling framework architecture for cross-layer resource orchestration, where service awareness provided by session control is effectively combined with the flexibility of software-defined network control. Following a hands-on approach, the proposed solution takes advantage of existing and commonly deployed technologies aiming at an incremental deployment of the software-defined control mechanisms for the purpose of cross-functional service orchestration. The signaling framework is presented and validated through experimental activities carried out on a test-bed reproducing a realistic cloud-based service scenario. Results are compared against an analytical model that allows to investigate and quantify the sensitivity of the cloud-based service performance to the most relevant system parameters. The study demonstrates that a critical role is played by the limitations of the response time of real devices such as commercial routers, and highlights how orchestration functions can mitigate the effect of such limitations while addressing scalability of the overall system. Cerroni, Walter; Gharbaoui, Molka; Martini, Barbara; Campi, Aldo; Castoldi, Piero; Callegati, Franco Cerroni, Walter; Gharbaoui, Molka; Martini, Barbara; Campi, Aldo; Castoldi, Piero; Callegati, Franco