Carlos Raniery Paula dos Santos

Management requirements for ClickOS-based Network Function Virtualization

Network Functions Virtualization (NFV) is a new approach to design, deploy, and manage network functions. In a recent past, such functions used to be implemented at hardware. This approach, besides effective, presents many disadvantages such as increased operational costs, difficulties to scale up or down the network, and deploy new functions. The rise of virtualization technologies, on the other side, provides new ways to rethink about network functions. Instead of specialized and expensive hardware, multiple network functions can share the same commodity hardware, thus contributing to a better utilization of resources. Besides its advantages, NFV is still on its early stages of employment. Important aspects are not yet being investigated by the research community. For example, to this date, the management requirements of NFV remain unclear. Therefore, the present paper addresses this subject, it presents a realistic network function request, which is used to identify management requirements in the context of a specific NFV enabler platform called ClickOS.

On the feasibility of Web 2.0 technologies for network management: A mashup-based approach

Mashups are a new breed of Web applications, created through the integration of external resources available on the Web. Recently, they have been considered a hallmark of Web 2.0 technologies, placing the end user on a developer role and encouraging both collaboration and reuse. Following the increasing efforts in investigating new approaches to network management, mashups present themselves as a technology that can bring several advantages to the field. However, to this date, the usage of mashups in network management remains unexplored. Therefore, the present paper approaches this subject, proposing a Mashup Development Tool to network management. We discuss both the architecture of such system and a proof of concept prototype. We them employ our prototype to address the case study of integrating Autonomous System routing information.

On using mashups for composing network management applications

Mashups are web applications created through the integration of external resources available on the web. They have been considered a hallmark of Web 2.0 technologies, allowing end users to develop their own applications and encouraging cooperation and reuse. However, their usage in the network management field remains unexploited. In this context, we look at Web 2.0 as a feasible mechanism able to integrate heterogeneous management information. In this article, we propose an architecture and a system prototype that allows network administrators to design their own management applications through the composition of external resources. The creation of mashups for two network management scenarios allowed us to both evaluate our architecture and, mainly, observe benefits, challenges, and opportunities that arise from such an approach. Observed characteristics, such as ease of use, extensibility, and contextspecific development, make mashups a candidate for an interesting area for further research in the network management field.

Monitoring Virtual Nodes using mashups

Abstract The use of virtualization technologies is one of major trends in computer networks. Up to now, most of monitoring tasks on Virtual Nodes, made up of several system virtualization environments and network virtualization environments, require manual intervention via non-standardized interfaces. Although monitoring based on proprietary command lines and graphical user interfaces may be enough for homogeneous Virtual Nodes, it is certainly not suitable for monitoring, in an integrated way, Virtual Nodes in which, the aforementioned environments use heterogeneous virtualization technologies, both in networks and systems. In this paper, we demonstrate that the mashup technology can be used to carry out the integrated monitoring of heterogeneous Virtual Nodes. In this sense, we present a mashup-based architecture targeted to monitor such type of Virtual Nodes, we introduce a reference implementation of the mashup-based architecture, and we develop on it, three monitoring mashups. The quantitative assessment of these mashups corroborates that they generate low traffic and have short response time. Furthermore, their qualitative assessment reveals that it is feasible to provide flexible and extensible mashups for monitoring Virtual Nodes.

Botnet master detection using a mashup-based approach

Botnets are considered by specialists, in both industry and academy, as one of the greatest threats to security on the Internet. These networks are composed by a large number of malware-infected hosts acting under a central command. They are usually employed to perform DDoS attacks or phishing scams. The behaviour of these botnets evolves due the adoption of new and sophisticated infection methods, changing of network protocols, and the employment of different command and control mechanisms. The security community, thus, is always dealing with such constant change. However, most botnet mitigation methods address just specific infection types or C&C protocols. We, therefore, propose a botnet mitigation approach based on the dynamic integration of pre-existing tools that can be employed together to achieve a more efficiently detection solution. To such end, we base our approach on a novel Web 2.0 technology called mashups to perform the information correlation. The proposal is extensible enough to allow even non-security information such as online mapping APIs be integrated to create more sophisticated compositions, and displaying the results in a more meaningful way.

On the design and performance evaluation of notification support for P2P-based network management

In this paper we present the design of a notification service used in a peer-to-peer (P2P) based network management solution called ManP2P. The notification service is based on the publish/subscribe paradigm, and implemented over a P2P overlay that carry the notification messages using SOAP, the Web services basic protocol. The performance of the notification service is evaluated considering the network traffic, the delivery delay, and the processing power required to forward notifications from physical devices (e.g., router, switches) to interested network managers. The results show that the notification service performs better if an increasing number of intermediate elements called mid-level managers is used between a notification source and destination.

Comparing virtualization solutions for NFV deployment: A network management perspective

Network Functions Virtualization (NFV) is a paradigm designed to promote service agility and able to quickly generate revenue, thus encouraging competition among companies in the computer network industry. Besides the advocated benefits of NFV, management requirements should be properly taken into account. The choice of a particular NFV-based technology must consider its management requirements. However, there is still no evaluation of virtualization solutions providing an in-depth analysis from the management point-of-view. This paper presents a performance analysis of three prominent virtualization solutions: ClickOS, CoreOS, and OSv. Our results place ClickOS and CoreOS as the best solutions regarding boot time, response time, and memory consumption. Moreover, based on the results obtained for each performance metric, we provide a broad discussion about the effectiveness of each virtualization solution in fulfilling qualitative management requirements.

The interplay between timeliness and scalability in cloud monitoring systems

Cloud computing is a groundbreaking solution to acquire computational resources on demand. To deliver high quality cloud services and provide features such as reduced costs and availability to customers, a cloud, like any other computational system, needs to be properly managed in accordance with its characteristics (e.g., scalability, elasticity, timeliness). In this scenario, cloud monitoring is a key to achieve it. To properly work, cloud monitoring systems need to meet several requirements such as scalability, accuracy, and timeliness. This paper aims to unveil the trade-off between timeliness and scalability. Evaluations demonstrate the mutual influence between scalability and timeliness based on monitoring parameters (e.g., monitoring topologies, frequency sampling). Results show that non-deep monitoring topologies and decreasing the frequency sampling assist to reduce the mutual influence between timeliness and scalability.

Refactoring internet of things middleware through software-defined network

Refactoring means to rewrite pieces of code aiming to improve it internally but keeping the expected software behavior. In this paper we present the refactoring of an Internet of Things middleware based on Software-Defined Network. In a previous work we proposed a middleware to address issues we found in healthcare devices used to monitor patients with chronic illnesses in their homes. Software-Defined Network allowed the redesign of the middleware architecture to improve things management, its interconnection with services, and the deployment process of new monitoring scenarios. Refactoring process also extended the middleware to support multiple services in a single home network sharing the same network infrastructure. This work details an OpenFlow controller and an application developed to achieve our goals; we also present sample scenarios where our approach can be applied showing different services delivered in the same home network environment, and using data from all connected devices to build a digital representation of the physical realm.

Taxonomy for the Network and Service Management Research Field

AbstractNetwork and service management has established itself as a research field in the general area of computer networks . However, up to now, no appropriate organization of the field has been carried out in terms of a comprehensive list of terms and topics. In this paper, we introduce a taxonomy for network and service management. With such a taxonomy, it is possible to better understand the landscape of research as well as to reason about possible future challenges and opportunities. As such, in addition to the taxonomy itself, we also present an initial analysis of the field’s past, present, and future, based on the records of papers submitted and accepted in major conferences in the area, as well as a site survey performed through a questionnaire answered by experts from both industry and academia.