Francisco José Monaco

Real-time compression of SOAP messages in a SOA environment

This paper discusses how the use of compression techniques aimed at decreasing data transfer times over a communication network can influence the response time of an application that process SOAP messages in the context of a service-oriented architecture. Following an overview of the most known object models and comparing some of their features, the article presents an heuristic that can be used to decide whether a soap message either should or should not be compressed. A simulated experiment shows that the proposed heuristic can help in reducing the service response time in a variety of scenarios.

Remote device command and resource sharing over the Internet: a new approach based on a distributed layered architecture

In addition to the remote access and computer-augmented functionality brought about by the earliest modalities of distance operation, technical advances in the form of telematics have opened up a whole new range of applications, one of which, resource sharing, deserves special attention. Nonetheless, while distance operations through computer networks, and particularly over the Internet, have attracted a great deal of attention in recent years, there is still a noticeable lack of important acquisitions regarding the systemic treatment of essential issues in this field. This paper presents an overview of the current trends in this emerging interdisciplinary area and briefly comments on the fundamentals of telematics-supported distance operation. A case study is used to report on an experience involving methodological investigations in this area.

An orthogonal real-time scheduling architecture for responsiveness QoS requirements in SOA environments

This paper introduces a novel design approach for real-time multi-resource scheduling aimed at ensuring upper-bounded average response time in interactive computer services. By analyzing the properties of conventional heuristics, this work presents a new algorithm which combines resource and queue disciplines in an orthogonal scheduling architecture which extends recent results on stochastic real-time operations with applications on SOA and pervasive computing environments. Simulation experiments shows how the attributes of the proposed technique can outperform those of conventionally employed alternatives.

Transient performance evaluation of cloud computing applications and dynamic resource control in large-scale distributed systems

This paper discusses on non-stationary performance evaluation and dynamic modeling of cloud computing environments. In computer systems, dynamic effects results from the filling of buffers, event-handling delays, non-deterministic I/O response times, network latency, among other factors. While computer systems performance evaluation under stationary workloads have met the needs of many engineering problems, new challenges arise as the deployment of increasingly complex and large-scale distributed systems becomes commonplace. One key aspect of this discussion is that transient analysis models how the system reacts to changes in the workload and may reveal that the resources necessary to support a high steady-state workload may not be sufficient to handle a small, but sudden, workload change, even of intensity far smaller than that supported by the systems stationary capacity. This article elaborates on these issues under a control-theoretical approach.

A novel QoS modeling approach for soft real-time systems with performance guarantees

This paper introduces a systematic approach for modeling QoS requirements of soft real-time systems with stochastic responsiveness guarantees. While deadline miss ratio and its proposed extensions have been considered for evaluating firm real-time systems, this work brings out its limitations for assessing the performance of emerging computer services operating over communication infrastructures with non-deterministic dynamics. This work explains how delay frequencies and delay lengths can be both represented into a single quantitative meaningful measure for performance evaluation of soft real-time constrained computer applications. It also explores the presented approach in the design of scheduling strategies which can ground novel business models for QoS-enabled service-oriented systems.

Restraining Content Explosion vs. Constraining Content Growth

As users learn new ways to explore the potential of Web, the information content in sites that are effectively explored tends to grow. Web technology is so powerful and flexible that the the limit of information management is usually reached before other technical restrictions and before we make the most of what Web can offer. In this paper we investigate the scenario where the content growth supersedes the capacity of information maintenance. We also present a strategy to deal with this problem without resorting to policies to constrain the growth of information content. By means of a case study we exemplify how a successful implementation was used to avoid the bottlenecks that contribute to content explosion.

Operating System from the Scratch: A Problem-based Learning Approach for the Emerging Demands on OS Development

Abstract In recent past history of computer systems industry, for decades, the hegemony of a few de facto standards dictated by major proprietary commercial products dominated the Operating Systems (OS) field. In such technological context, conso- nantly to this trend, the knowledge objective focused by academical and training courses on OS-related disciplines has often been addressed more from the stand point of essential theoretical background than of the technical skills for actuation on de- sign and development field. Emerging paradigms, nevertheless, have been rapidly changing this scenario. Among them, the establishment of Open Source concept is boosting the growing diversity of new operating systems; concomitantly, evolution of embedded hardware architectures has make it possible to run sophisticated operating systems where only bare rudimentary, ad hoc system-software were once practical. Aligned along this perspective, this paper introduces a new platform for teaching and training programs on OS development founded on a project-based approach which guides the student throughout the process of designing and programming a sufficiently simple, but yet realistic and fully functional, OS from the scratch. The differential of the present proposal regarding related works is that, instead of either merely inspecting example-code or experimenting with simulators, the apprentice is guided across the challenge of coding an entire new instance of a didactic system specification. A comparison of the companion OS-example with existing alternatives brings out a less complex implementation structure which maps conceptual modules with implementation blocks in an intuitive correspondence and with reduced function cou- pling. Moreover, the learning platform comes with a courseware material consistently linked to the laboratory practices, and aimed at the systemic comprehension of the many related multidisciplinary aspects.

Feedback-Based Adaptive Resource Control in QoS-Aware SOA Systems with Soft Real-Time Requirements

When deployed as operational components of production systems, novel computer services are supposed to respond synchronously to real-world events associated to the business process they implement, thereby needing to meet temporal constraints dictated by the dynamics of the environment in which they operate. This elicits a real-time system approach. One emerging concept to cope with such unpredictability of large-scale distributed computer applications is the use of feedback control principles. This paper introduces a feedback-based adaptive resource control algorithm for composite applications implementing real-time business process. The study is based on recent achievements in the field and ongoing progresses. A brief background on the field, the rationales of the proposed techniques and development results are presented.

Extending discrete-event simulation frameworks for non-stationary performance evaluation: requirements and case study

This paper introduces an approach to obtain an empiric analytical model of the performance dynamics of computer systems out of discrete-event simulation (DES) experiments. To that end, the proposed methodology elicits the requirements for extending conventional stationary DES frameworks so as to meet the needs of transient performance analysis. The work goes through the rationales for the conceptual formulation of dynamic (as opposed to static) capacity and summarizes a methodology for system identification. Results of on going research are reported and conclusions illustrated by case study.

Embedded Linux in Real-Time Applications: Performance Enhancements of Experimental Fully-Preemptible Capabilities over the Standard Kernel in a Critical Mobile System

This paper presents the results of a performance evaluation of a standard Linux-based control system employed in many critical embedded applications, and identify improvements in responsiveness and efficiency obtainable by specific modifications aimed at enhancing the preemptive and real-time capabilities over the standard OS kernel. The study was accomplished with the aid of a tool for runtime performance measurement developed in the scope of the INCT-SEC GT4 activities, which collects information of the interaction between the application, operating system and hardware during the process execution. Experiments carried out with the embedded control software of another running project of INCT-SEC, a tactical indoors robot, reveal potential significant improvements in efficiency and responsiveness which can be exploited by merely replacing the kernel, without any modification on the application itself. It is also offered an insight on how the characteristic of the application in terms of CPU and I/O usage patterns can influence the system performance, and how such knowledge can be used in the project of embedded application such as to optimize the dependability and efficiency of critical applications.