An overview of the self-organizing map algorithm, on which the papers in this issue are based, is presented in this article.

The Self-Organizing Map

コンピュータ・サイエンス : ACM computing surveys

Rainbow is a framework for engineering a system with run-time, self-adaptive capabilities to monitor, detect, decide, and act on opportunities for system improvement. We applied Rainbow to a system, Znn.com, and evaluated its effectiveness to self-adapt on three levels: its effectiveness to maintain quality attribute in the face of changing conditions, run-time overheads of adaptation, and the engineering effort to use it to add self-adaptive capabilities to Znn.com. We make Znn.com and the associated evaluation tools available to the community so that other researchers can use it to evaluate their own systems and the community can compare different systems. In this paper, we report on our evaluation experience, reflect on some principles for benchmarking self-adaptive systems, and discuss the suitability of our evaluation tools for this purpose.

/pdf/evaluating-the-effectiveness-of-the-rainbow-self-adaptive-1dj7y4b909.pdf

Evaluating the effectiveness of the Rainbow self-adaptive system

Modern, complex software systems (e-commerce, IT, critical infrastructures, etc.) are increasingly required to continue operation in the face of change, to self-adapt to accommodate shifting user priorities, resource variability, changing environments, and component failures. While manual oversight benefits from global problem contexts and flexible policies, human operators are costly and prone to error. Low-level, embedded mechanisms (exceptions, time-outs, etc.) are effective and timely for error recovery, but are local in scope to the point-of-failure, application-specific, and costly to modify when adaptation objectives change. An ideal solution leverages domain expertise, provides an end-to-end system perspective, adapts the system in a timely manner, and can be engineered cost-effectively.
Architecture-based self-adaptation closes the “loop of control,” using external mechanisms and the architecture model of the target system to adapt the system. An architecture model exposes important system properties and constraints, provides end-to-end problem contexts, and allows principled and automated adaptations. Existing architecture-based approaches specialize support for particular classes of systems and fixed sets of quality-of-service concerns; they are costly to develop for new systems and to evolve for new qualities. To overcome these limitations, we posit this thesis:
We can provide software engineers the ability to add and evolve self-adaptation capabilities cost-effectively, for a wide range of software systems, and for multiple objectives, by defining a self-adaptation framework that factors out common adaptation mechanisms and provides explicit customization points to tailor self-adaptation capabilities for particular classes of systems, for multiple quality-of-service objectives.
Our approach, embodied in a system called Rainbow, provides an engineering approach and a framework of mechanisms to monitor a target system and its environment, reflect observations into the system's architecture model, detect opportunities for improvements, select a course of action, and effect changes. The framework provides general and reusable infrastructures with well-defined customization points, a set of abstractions, and an adaptation engineering process, focusing engineers on adaptation concerns to systematically customize Rainbow to particular systems. To automate system self-adaptation, Rainbow provides a language, called Stitch, to represent routine human adaptation knowledge using a core set of adaptation concepts.

/pdf/rainbow-cost-effective-software-architecture-based-self-33gijan84m.pdf

Rainbow: cost-effective software architecture-based self-adaptation

Increasingly, systems must have the ability to self-adapt to meet changes in their execution environment. Unfortunately, existing solutions require human oversight, or are limited in the kinds of systems and the set of quality-of-service concerns they address. Our approach, embodied in a system called Rainbow, uses software architecture models and architectural styles to overcome existing limitations. It provides an engineering approach and a framework of mechanisms to monitor a target system and its environment, reflect observations into a system’s architecture model, detect opportunities for improvement, select a course of action, and effect changes in a closed loop. The framework provides general and reusable infrastructures with well-defined customization points, allowing engineers to systematically customize Rainbow to particular systems and concerns.

https://www.springer.com/cda/content/document/cda_downloaddocument/9780387898278-c2.pdf?SGWID=0-0-45-734916-p173871105

Software Architecture-Based Self-Adaptation

This work proposes using a neural network with self organizing maps, to build a configuration policy, which enables the management of a supporting infrastructure for Web applications using virtual machines. The neural network classifies the cluster´s operation states to perform configuration operations adding or subtracting resources from the cluster. The overall goal is to ensure quality of service required by the application, while trying to save energy, acting efficiently on physical servers (hosts) or manipulating the VMs. The work includes a performance evaluation carried out over a system implemented based on the proposed architecture.

Use of Neural Networks in the Management of Virtual Web Servers

O escalonamento de processos e um assunto que requerdos alunos dos cursos de graduacao em Ciencias daComputacao uma alta capacidade de abstracao. Seriadesejavel uma ferramenta que auxiliasse na compreensaodos aspectos cuja compreensao e mais dificil quandoapresentados apenas pelos metodos de ensino convencionais.Desenvolvemos um simulador de mecanismos dealocacao da CPU que se propoe a facilitar a tarefa deavaliar o impacto da implantacao de um determinadomecanismo de alocacao da CPU sobre um sistema computacional.O simulador fornece aos usuarios meios decriar mecanismos de alocacao da CPU de forma personalizadae, em seguida, testa-los sobre diferentes condicoesde funcionamento. Conseguimos um software didatico,extensivel, reutilizavel, modular e independente deplataforma.

/pdf/implementando-um-simulador-de-mecanismos-de-alocacao-da-cpu-34ieiins2q.pdf

Implementando um Simulador de Mecanismos de Alocação da CPU

This paper presents the DistributedCL, a framework that provides the applications developed using the OpenCL interface location-transparent GPU processing. The application can explore distributed processing with no modification. The architecture of the framework and the programming model are presented, and the possible performance bottlenecks are discussed.

DistributedCL: A Framework for Transparent Distributed GPU Processing Using the OpenCL API

A Virtualizacao de Funcoes de Rede aumenta a flexibilidade e reduz custos de infraestruturas de telecomunicacoes. Para tal, as funcoes de rede sao implementadas em servidores virtualizados, ao inves de embarcadas em equipamentos dedicados. Um dos desafios dessa abordagem e o posicionamento de funcoes virtuais de rede, que deve ser resiliente e energeticamente eficiente. Este trabalho formula um problema de programacao inteira mista para posicionar tais funcoes, escolhendo os servidores para atenderas demandas de servico. O problema formulado minimiza o uso da energia e prove resiliencia aos servicos atraves de replicacao de funcoes. Como solucao, propoe-se a heuristica TRELIS, que reduz o tamanho do problema e economiza ate 15% da energia.

/pdf/trelis-posicionamento-de-funcoes-virtuais-de-rede-com-4qlk13fk6x.pdf

TRELIS: Posicionamento de Funções Virtuais de Rede com Economia de Energia e Resiliência

Os sumarios medicos disponibilizados na forma de ferramentas digitais tem sido cada vez mais usados para apoiar a tomada de decisao medica. Apesar de amplamente disponiveis, tais ferramentas apresentam barreiras para sua adocao, entre elas o custo elevado. Este trabalho apresenta a proposta de uma plataforma de sumarios medicos utilizando o modelo colaborativo, crowdsourcing, como forma de superar as barreiras das ferramentas tradicionais, onde os proprios usuarios sao responsaveis por adicionar e manter os resumos. O desenvolvimento da plataforma, que inclui aplicativos web e mobile encontra-se em andamento.

/pdf/proposta-de-uma-plataforma-crowdsourcing-para-apoio-a-tomada-l2ra7fajef.pdf

Alexandre Sztajnberg

Papers

Use of Neural Networks in the Management of Virtual Web Servers

Implementando um Simulador de Mecanismos de Alocação da CPU

DistributedCL: A Framework for Transparent Distributed GPU Processing Using the OpenCL API

TRELIS: Posicionamento de Funções Virtuais de Rede com Economia de Energia e Resiliência

Proposta de uma Plataforma Crowdsourcing para Apoio à Tomada de Decisão Médica