Antonio de Barros Serra

Assuring QoS differentiation and load balancing on web servers clusters

The number of users of services deployed on the Internet has been increasing continually. This has been causing overload in a great amount of Web servers reducing QoS (quality of service) offered by service providers. Furthermore, users with different QoS aspirations have been forced to share available processing resources in a same way even they can pay more. To deal with these problems, a possible solution is the balancing of imposed load among a certain number of computers to improve performance and the use of admission control mechanisms to allow differentiated allocation of resources for specific service classes. This paper presents WS-DSAC (Web servers ifferentiated services admission control), a diffserv-based admission control and load balancing mechanism conceived to improve QoS on Web servers clusters. WS-DSAC has three main aims: to balance the imposed load, to guarantee different QoS levels and to use available resources in an effective way. We also present evaluation experiments showing the mechanism is able to achieve these objectives

SPEWS: A Framework for the Performance Analysis of Web Services Orchestrated with BPEL4WS

This paper addresses quality of service aspects of Web Services (WS) orchestrations created using the Business Process Execution Language for Web Services (BPEL4WS). BPEL4WS is a promising language describing the WS orchestrations in form of Business Processes, but it lacks of a sound formal semantic, which hinders the formal analysis and verification of business processes specified in it. Formal methods, like Petri Nets (PN), may provide a means to analyse BPEL4WS processes, evaluating its performance, detecting weaknesses and errors in the process model already at design-time. A framework for transformation of BPEL4WS into Generalized Stochastic Petri Nets (GSPN) is proposed to analyse the performance and throughput of WS, based on the execution of orchestrated processes.

A load-balancing distributed platform based on differentiated services for a telecare application

Telecare applications are composed by different kinds of services. Some of these services require high level availability and performance because they involve human lives. One possible solution to improve availability and performance in these applications is to make an efficient distribution of service avoiding concentration of several tasks on only one server. This work presents a load-balancing platform based on differentiation of services that increases application-level QoS (quality of service) in a telecare application. The platform allows the coexistence of Internet services and distributed objects. We also present experimental results showing that the platform balances the imposed load and differentiates the QoS offered to different services.

DynaMOC: a dynamic overlapping coalition-based multiagent system for coordination of mobile ad hoc devices

In this work, we focus on problems modeled as a set of tasks to be scheduled and accomplished by mobile autonomous devices that communicate via a mobile ad hoc network. In such situations, the communication cost, computational efforts and environment uncertainty are key challenges. It is intuitive to consider that keeping information about tasks globally known by devices can provide better schedules. However, there are some contexts - such as those where tasks require startup based on location - where information restricted to coalitions of devices can still produce satisfactory scheduling. The existing heuristics, however, do not consider this approach. In this paper, we propose a multiagent system that coordinates the dynamic formation of overlapping coalitions and the scheduling of tasks within them. Heuristics for calculating the size of coalitions, as well as for scheduling tasks are proposed based on a Markov decision process. The system is applied to solve the problem of area coverage in a simulated environment and the results show that good schedules are obtained with lower cost of communication and computation compared with the solution based on globally known information.

DynaMOC: A Multiagent System for Coordination of Mobile Teams Based on Dynamic Overlapping Coalitions

In this work, we focus on problems modeled as a set of activities to be scheduled and accomplished by mobile autonomous devices that communicate via a mobile ad hoc network. In such situations, the communication cost and computational efforts are key challenges. It is well known that keeping information about activities globally known by devices can provide better schedules. However, there are some contexts such as those where activities require startup based on location, whereby information restricted to neighborhood coalitions of devices can still produce satisfactory scheduling. The existing heuristics do not consider this approach. In this paper, we propose a multi-agent system that coordinates the dynamic formation of overlapping coalitions and the scheduling of activities within them. Heuristics for calculating the size of coalitions, as well as for scheduling activities, are proposed. The system is applied to solve the problem of area coverage in a simulated environment, whereby the results show that satisfactory schedules are obtained with lower cost of communication and computation in comparison with the solution based on globally known information.

Model Method for the Transformation of BPEL4People into Generalized Stochastic Petri Nets

WS-BPEL Extension for People (BPEL4People) introduces human activity to Web Services Business Process Execution Language (WSBPEL/ BPEL). Some works have been done on the performance verification of WS-BPEL, but there are not works on the performance verification of WSBPEL4People. In this paper, we propose a model method for the transformation of WS-BPEL4People / BPEL4People into Generalized Stochastic Petri Nets (GSPN) and this way evaluate its performance. During the transformation, Petri Net is used to model BPEL activities and the human activity of a human task. Then, by validating the generated model, some potential problems with performance of WS-BPEL4People is detected.

A Semantic Framework for Resource Discovery Based on Ontology

In this paper, we propose a better solution to the limitations involved by using the current standard of syntactic matching for resource discovery in Grid environment. Our observations noted that syntactic matching was limited in its ability to use all available resources. In order to overcome these limitations, we hereby propose a method for resource discovery involving semantic matching based in ontology and explain the industry’s available tools that were involved in implementing this structure. After clearly presenting all aspects of the structure and how to implement it, we offer the results from various simulations, which clearly demonstrate the greater results obtained through semantic matchmaking. In conclusion, we take a brief look at some of the future work that will be possible as a result of our findings.

The Simulation of DARC (Dynamic Architecture for Reconfiguration of Web Servers Clusters) in Petri Nets

Allocating resources in a web server cluster is usually performed nowadays by a cluster administrator. However, due to the dynamics of the Internet as far as the use of resources is concerned, such a task may be considered critical and inefficient if accomplished manually. This paper presents an architecture that performs an automatic dynamic reconfiguration using a multi agent system. The architecture, a formal specification in Coloured Petri nets and experimental results of the implementation of the solution are presented.