Christian Esposito

Survey On reliability in publish/subscribe services

Modern large-scale mission-critical systems demand efficient and robust multi-point data dissemination infrastructures. Since such infrastructures have to exhibit good performance when scaling up the number of interacting entities and managing large amounts of data, publish/subscribe services represent a suitable middleware solution due to their decoupling properties. However, since data are conveyed by networks where failures may occur, and since nodes may present a faulty behavior, such services also have to adopt proper mechanisms to deal with several kinds of failures and to guarantee event dissemination despite their occurrence. Although significant efforts have been made on this topic, many issues are still open. This article covers an introduction to the principles of assuring event notification even in the presence of faults, and an analysis of relevant state-of-the-art by both surveying the academic literature over the period 2000-2011 on reliable publish/subscribe services and overviewing current marketed products. Then, it presents the main challenges that are still unresolved and are worth being addressed in future research efforts.

Reliable publish/subscribe middleware for time-sensitive internet-scale applications

Federating mission critical systems over wide-area networks still represents a challenging issue. For example, it is hard to assure both reliability and timeliness in a hostile environment such as Internet. The publish/subscribe (pub/sub) interaction model is a promising solution for scalable data dissemination over wide-area networks. Nevertheless, currently available pub/sub systems lack efficient support to achieve both reliability and timeliness in unreliable scenarios. This paper describes an innovative approach to fill this gap making three contributions. First, a cluster-based peer-to-peer organization is introduced to handle a large number of publishers and subscribers. Second, the cluster coordinator is replicated to mask process crashes and to preserve cluster connectivity toward the outside world. Third, multiple-tree redundancy is applied to tolerate link crashes thereby minimizing unpredictability in the delivery time. We present a simulation-based evaluation to assess the effectiveness of our approach in an unreliable setting. This study indicates that our approach enforces the reliability of event delivery without affecting its timeliness.

Calibrating Indoor Positioning Systemswith Low Efforts

Recently, the positioning techniques based on the IEEE 802.11 signal strength are becoming the dominant solutions in the mobile device localization within indoor scenarios. Such solutions are characterized by two main pitfalls that compromise their effective usage in real application environments. First, during the calibration, a large amount of manual effort is required for acquiring a massive collection of training samples. Second, the positioning accuracy is directly related to the deployment of the wireless access points into the workspace, which is extremely time-consuming and requires human intervention. This paper presents an approach to reduce the manual calibration and to optimize the positioning accuracy, by selecting the best deployment schema of the wireless access points. The approach has been implemented in a tool, which uses an analytical signal propagation model to build the radio map of a given workspace, and exploits a multi-objective genetic algorithm to identify the best access points placement pattern that fits the required accuracy. A detailed experimental campaign is presented in order to show the benefits achievable by the proposed approach.

Performance assessment of OMG compliant data distribution middleware

Event-driven architectures (EDAs) are widely used to make distributed mission critical software systems more- efficient and scalable. In the context of EDAs, data distribution service (DDS) is a recent standard by the object management group that offers a rich support for quality- of-service and balances predictable behavior and implementation efficiency. The DDS specification does not outline how messages are delivered, so several architectures are nowadays available. This paper focuses on performance assessment of OMG DDS-compliant middleware technologies. It provides three contributions to the study of evaluating the performance of DDS implementations: 1) describe the challenges to be addressed; 2) propose possible solutions; 3) define a representative workload scenario for evaluating the performance and scalability of DDS platforms. At the end of the paper, a case study of DDS performance assessment, performed with the proposed benchmark, is presented.

Challenges in Delivering Software in the Cloud as Microservices

Microservices can be broadly defined as the design of service-oriented software using a set of small services. In a microservice architecture, application complexity is distributed among narrowly focused and independently deployable units of computation. Such complexity can result in security vulnerabilities. Trustworthiness is also an issue when dealing with microservices. Moreover, there may be gaps in existing legal frameworks with regard to this technology. Solutions to these issues must seek balance between security and performance.

Deployment of RSS-Based Indoor Positioning Systems

Location estimation based on Received Signal Strength (RSS) is the prevalent method in indoor positioning. For such positioning systems, a massive collection of training samples is needed for their calibration. The accuracy of these methods is directly related to the placement of the reference points and the radio map used to compute the device location. Traditionally, deploying the reference points and building the radio map require human intervention and are extremely time-consuming. In this paper we present an approach to reduce the manual calibration efforts needed to deploy an RSS-based localization system, both when using only one RF technology or when using a combination of RF technologies. It is an automatic approach both to build a radio map in a given workspace by means of a signal propagation model, and to assess the system calibration that best fits the required accuracy by using a multi-objective genetic algorithm.

On data dissemination for large-scale complex critical infrastructures

Middleware plays a key role for the achievement of the mission of future large scale complex critical infrastructures, envisioned as federations of several heterogeneous systems over Internet. However, available approaches for data dissemination result still inadequate, since they are unable to scale and to jointly assure given QoS properties. In addition, the best-effort delivery strategy of Internet and the occurrence of node failures further exacerbate the correct and timely delivery of data, if the middleware is not equipped with means for tolerating such failures. This paper presents a peer-to-peer approach for resilient and scalable data dissemination over large-scale complex critical infrastructures. The approach is based on the adoption of epidemic dissemination algorithms between peer groups, combined with the semi-active replication of group leaders to tolerate failures and assure the resilient delivery of data, despite the increasing scale and heterogeneity of the federated system. The effectiveness of the approach is shown by means of extensive simulation experiments, based on Stochastic Activity Networks.

Architectural Evolution of Monitor and Control Systems - Issues and Challenges

Large-scale Complex Critical Infrastructures (LCCIs), such as power grids and transport infrastructures ( e.g. , network of airports and seaports), play a key role into several fundamental human activities, and represent the next generation of Monitor and Control Systems. They make extensive usage of Information and Communications Technology ( e.g. , computing systems, communication networks, and sensing hardware) for providing support for advanced monitoring and control facilities. However, solutions currently adopted in the field exhibit several inefficiencies when applied to LCCI. On the contrary, Distributed Event Based Systems (DEBS) seam promising solutions due to their intrinsic decoupling properties that enforce strong scalability degrees. However, they also present some open issues when dealing with the challenges imposed by LCCIs. This paper provides an introduction on LCCIs, a presentation of their challenges, and an analysis on the adoption of DEBS for LCCIs.

Live Migration in Emerging Cloud Paradigms

The elastic provisioning of resources and the capability to adapt to changing resource demand and environmental conditions on-the-fly are, probably, key success factors of cloud computing. Live migration of virtual resources is of pivotal importance in achieving such key properties. However, the ability to effectively and efficiently determine which resource to be migrated and where, by satisfying proper objectives and constraints, remains a research challenge. The existing literature is generally based on metaheuristics running a central resolver. Such an approach is not suitable because it only considers the quality-of-service aspect during the decision-making performance while ignoring the regulatory challenges. This column highlights the regulatory challenges associated with the cross-border dataflow implication of migration and stresses the need to adopt alternative decision approaches.

Achieving reliable and timely event dissemination over WAN

The design of large-scale critical infrastructures demands for innovative data dissemination services, able to jointly provide reliability and timeliness guarantees. Current middleware solutions do not address both these aspects. Indeed, fault tolerance is typically achieved at the cost of severe performance fluctuations, or timeliness is always obtained by softening the fault-tolerance requirements. In this paper we propose to fulfill this lack by combining two different approaches, namely coding and gossiping. We provide a theoretical model to evaluate the potential benefit of coding on the information delivery performance. These results are also confirmed by an experimental analysis conducted on a real air traffic control workload, which evidences how coding mitigates latency and overhead penalties to ensure reliable event notification.