Marcelo Anderson Santos

Dependability assessment of virtualized networks

One of the key characteristics of virtual networks (VN) is their intrinsic dynamical aspects. VNs allow network operators to have on-demand negotiation of a variety of services. Risks are inherent to virtualized infrastructures since the underlying physical network components are failure-prone. Therefore evaluation and analysis of risks in VNs, from assessment of dependability attributes, are of paramount importance for accurate management and control tasks. This paper proposes and evaluates a method to assess dependability in virtualized network environments. Through the use of Reliability Block Diagram (RBD) and Stochastic Petri Nets (SPN), we provide a first look at risks in network virtualization by means of dependability modeling and analysis. Our methodology allows the networking research community to develop new heuristics when dealing with resource allocation in VNs, while taking into account dependability attributes.

A Software Engineering Perspective on SDN Programmability

Software-defined networking (SDN) has received a great deal of attention from both academia and industry in recent years. Studies on SDN have brought a number of interesting technical discussions on network architecture design, along with scientific contributions. Researchers, network operators, and vendors are trying to establish new standards and provide guidelines for proper implementation and deployment of such novel approach. It is clear that many of these research efforts have been made in the southbound of the SDN architecture, while the northbound interface still needs improvements. By focusing in the SDN northbound, this paper surveys the body of knowledge and discusses the challenges for developing SDN software. We investigate the existing solutions and identify trends and challenges on programming for SDN environments. We also discuss future developments on techniques, specifications, and methodologies for programmable networks, with the orthogonal view from the software engineering discipline.

How Software Aging affects SDN: A view on the controllers

Software-Defined Networking proposes a new paradigm by separating the control plane from the data plane. The control planes main responsibility is deployed in a centralized component called SDN controller. Therefore, the SDN based network becomes dependent on a single functional entity (although it may have multiple instances) in order to accomplish its routing and policy rules. The SDN controller is vulnerable to degradation due to a well-known phenomenon called Software Aging. To the best of our knowledge, this work is the first that assesses the impacts of software aging on current implementation of SDN controllers. We show that such a phenomenon must not be neglected in future software architecture designs and implementations.

Model-driven networking: A novel approach for SDN applications development

Software-Defined Networking (SDN) has been receiving a great deal of attention from both academic and industry communities. One reason to this interest is that SDN enables the network programmability, through an external controller, which supports applications and policies built from SDN programming languages, thus breaking the traditional bind between control and data plane. Nevertheless, the application development in this context is still complex for such recent technology. Moreover, there is a strong need for methodologies and tools that explore the abstraction levels potentials supported by SDN. This paper presents a new approach based on the Model-Driven Engineering (MDE) paradigm, called Model-Driven Networking (MDN). MDN relies on a Domain-Specific Modelling Language (DSML) to create SDN applications. We argue that MDN raises the level of abstraction on development, thus reducing the complexity to implement SDN applications and avoiding inconsistent policies. In order to show the relevance and the technological viability of our proposal, we have specified a DSML and have built a tool for creating SDN applications using the MDN approach.

Role-Based Self-Appointment for Autonomic Management of Resources

Resource management in distributed environments, such as Cloud virtualized networks, is one of the main concerns of infrastructure providers. Despite many advantages of centralized solutions (such as the facility of implementation and management), their unique point of failure and the lack of scalability make us consider the use of distributed and autonomic solutions. Beyond this, centralized solutions would be inefficient to deal with dynamic and peak situations. By using simple sets of rules, each autonomic node can independently act and react according to scenarios changes without the knowledge of the whole system status, making autonomic management robust and adaptable. The main goal of this paper is to describe a role-based autonomic mechanism - called Role-Based Self-Appointment (RBSA) - for resource management in distributed environments and its applicability in different scenarios.

On the load balancing of virtual networks in distributed clouds

The distribution of computing resources in different geographical regions and the promotion of full integration with network resources are important issues of new architectures for Cloud computing. Such scattered Cloud deployments, called Distributed Clouds (D-Clouds), can directly reach users due to their inherently distributed infrastructure and the ownership of the network. Thus, D-Clouds can comply with geographically-based requirements and network-based quality of service. One of the challenges in this area is the resource management. In this way, a clever resource allocation algorithm is needed to satisfy service requirements and an owners management objectives. This paper proposes algorithms for allocation of computing and network resources in a D-Cloud with the objectives of balancing the load in the virtualized infrastructure and of considering constraints, such as processing power, memory, storage, and network delay. The evaluation of the algorithm shows that it is indeed adequate for link allocation across different physical networks.

Cooperative dynamic eHealth service placement in Mobile Cloud Computing

In the area of mobile applications, emerging eHealth applications used in conjunction with wearable medical sensor devices and personal devices are being adopted increasingly by people with the aim to improve their lifestyle and health. eHealth providers, willing to provide remote eHealth management, are integrating Wireless Body Area Networks (WBANs) technology and Cloud Computing. Cloud computing has the benefit to eliminate the need of maintaining costly hardware, software and network infrastructures. This combination of Cloud Computing and Mobile Computing is known as Mobile Cloud Computing (MCC). It allows eHealth providers to deploy instantly and on demand their eHealth services to monitor peoples health status. In this paper, we propose a cooperative strategy between a mobile operator and a cloud provider for an efficient eHealth service placement in the cloud which take into account the network latency. We highlight the importance of mutual coordination between them to minimize the cost of deployed services while maintaining customers satisfaction.

Towards Efficient Automatic Scaling and Adaptive Cost-Optimized eHealth Services in Cloud

Cloud Computing is an emerging commercial model which allows organizations to eliminate the need to maintain costly hardware, software and network infrastructures. It also permits to avoid the high operational cost for operating and maintaining these infrastructures. Similarly, in the eHealth area, emerging eHealth applications used in conjunction with wearable medical sensor devices and personal devices are being adopted by more and more people with the aim to improve their lifestyle and health. eHealth organizations, willing to provide remote eHealth management, are integrating Wireless Body Area Networks (WBANs) technology and Cloud Computing technology. This integration allows eHealth organizations to deploy their eHealth services on demand and instantly to monitor patients health status. We propose in this paper, a solution for such organizations to efficiently deploy their eHealth services and adapt provisioned physical resources dynamically to satisfy the quality of health of potentially millions of subscribers

Conducting Network Research in Large-Scale Platforms: Avoiding Pitfalls in PlanetLab

Performing distributed experimentation in an Internet-like scale is essential for the introduction of innovative protocols, applications, and techniques in future networks. In the past, having access to different servers in different locations in the world used to be a very hard task. This scenario has changed in recent years with the introduction of Planet Lab, a distributed laboratory with 1000+ servers spread in the five continents for developing experiments and measurements in a planetary scale. However, Planet Lab currently has some limitations that sometimes prevent researchers to perform reliable experiments. This paper presents a large-scale measurement study in Planet Lab aiming at shedding some light into the availability of resources for performing experimental research on it. Results show that users should take extra care when designing and deploying experiments in Planet Lab, due to high instability and heterogeneity of the available resources.

High-level modeling and application validation for SDN

Software-Defined Networking (SDN) enables applications running on its control plane. The Northbound API allows programmers to develop SDN applications for a number of policy-based network management tasks. However, there is still a clear need for supporting the development of controller-agnostic modeled applications. In this paper, we show how the Model-Driven Networking (MDN), a framework composed of CASE tool and Domain-Specific Modeling Language (DSML), can be a feasible solution to create applications independent from controllers and to enable proper verification of SDN applications. Our evaluation demonstrates that MDN framework is viable for using in real scenarios and independent from SDN controllers. Moreover, our performance tests show that: (i) MDNs code generation is two times faster than other approaches; and (ii) it can validate several constraints and complex topologies at millisecond-timescale.