Daniel Sadoc Menasche | Universidade Federal do Rio de Janeiro (UFRJ) (original) (raw)

Papers by Daniel Sadoc Menasche

ACM SIGMETRICS Performance Evaluation Review, 2020

The heavy traffic regime is a regime wherein system resources are always busy. As caches and soci... more The heavy traffic regime is a regime wherein system resources are always busy. As caches and social network timelines are intrinsically always busy, i.e., their space-shared resources are always utilized, the goal of this paper is to evaluate the implications of a simple albeit unexplored observation: caches and timelines operate under heavy traffic. First, we introduce the control problem of caches and timelines under heavy traffic. Then, we derive properties of the asymptotically optimal (AO) policy. In particular, we indicate that there is an AO control that threatens content diversity, as it involves maintaining contents from up to two classes in the system, leading to the so called filter bubbles.

2019 IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW), 2019

2017 13th Annual Conference on Wireless On-demand Network Systems and Services (WONS), 2017

Abstract. We consider publishers of large number of movies and music files. Such items, associate... more Abstract. We consider publishers of large number of movies and music files. Such items, associated to small costs to be replicated, are referred to as information goods. The items are produced by content providers, disseminated by publishers and consumed by users. Among the items offered in content provider's catalogs, the publishers must select which ones to buy and disseminate. The selection problem is referred to as the problem of catalog selection of information goods. The catalog selection is performed so as to maximize the ...

ACM SIGMETRICS Performance Evaluation Review, 2016

IEEE Transactions on Mobile Computing, 2020

2019 IEEE 27th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS), 2019

2020 IEEE 9th International Conference on Cloud Networking (CloudNet), 2020

Despite the tremendous success of SDNs in datacen-ters, their wide adoption still poses a key cha... more Despite the tremendous success of SDNs in datacen-ters, their wide adoption still poses a key challenge: the packet-forwarding rules in switches require fast and power-hungry memories. Rule tables, which serve as caches, are of limited size in cheap and energy-constrained devices, motivating novel solutions to achieve high hit rates. In this paper, we leverage device connectivity in the fast data plane, where delays are in the order of few milliseconds, and propose multiple switches to work together to avoid accessing the control plane, where delays are orders of magnitude greater. As a low priority rule in a cache entails caching higher priority rules, we pose the problem of cooperative caching with dependencies. We provide models and algorithms for cooperative rule caching with dependencies, accounting for dependencies among rules implied by existing switch memory types. We develop caching algorithms for several typical use cases and study the difficulty to find an optimal cooperative rule placement as a function of the matching pattern, which lay the foundations of cooperative caching with dependencies.

The demand for Internet services that require frequent updates through small messages, such as mi... more The demand for Internet services that require frequent updates through small messages, such as microblogging, has tremendously grown in the past few years. Although the use of such applications by domestic users is usually free, their access from mobile devices is subject to fees and consumes energy from limited batteries. If a user activates his mobile device and is in range of a service provider, a content update is received at the expense of monetary and energy costs. Thus, users face a tradeoff between such costs and their messages aging. The goal of this paper is to show how to cope with such a tradeoff, by devising aging control policies. An aging control policy consists of deciding, based on the current utility of the last message received, whether to activate the mobile device, and if so, which technology to use (WiFi or 3G). We present a model that yields the optimal aging control policy. Our model is based on a Markov Decision Process in which states correspond to message ...

2020 IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW), 2020

This extended abstract summarizes the background, goals, applicability domain, method, results, a... more This extended abstract summarizes the background, goals, applicability domain, method, results, and lessons learned presented in the corresponding chapter of the Handbook of Software Aging and Rejuvenation.

Anais XXXVIII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC 2020), 2020

O cubo mágico, também conhecido como cubo de Rubik, é um jogo popular que recentemente atraiu a a... more O cubo mágico, também conhecido como cubo de Rubik, é um jogo popular que recentemente atraiu a atenção da comunidade científica como um problema estilizado para ilustrar a aplicabilidade das técnicas de aprendizado de máquina. Neste artigo, mostramos novos resultados que aproveitam as propriedades de simetria do cubo Rubik para fins de roteamento. Dados dois estados alcançáveis do cubo, mostramos que podemos rotear eficientemente de um estado para outro, dada uma solução para o problema padrão do cubo Rubik. Em seguida, indicamos como o algoritmo eficiente proposto pode ser usado para refinar as soluções abaixo do ideal para o roteamento de caminho mais curto entre os estados do cubo Rubik.

iSys - Brazilian Journal of Information Systems, 2019

O Facebook tem um impacto significativo na vida de milhões de usuários da Internet, todos os dias... more O Facebook tem um impacto significativo na vida de milhões de usuários da Internet, todos os dias. Entretanto, os mecanismos usados pelo Facebook para filtrar mensagens apresentadas aos usuários não são de domínio público, motivando uma engenharia reversa dos mesmos. Neste trabalho, propomos modelos e medições para melhor entender o comportamento de timelines. Em particular, reportamos resultados sobre medições de visibilidade de fontes das eleições italianas de 2018, que motivam um modelo analítico para caracterizar a visibilidade de posts. Dentre as implicações práticas de nossos estudos, indicamos seu potencial para inferir diferentes métricas de visibilidade a partir de medidas colhidas do sistema.

2020 32nd International Teletraffic Congress (ITC 32), 2020

IEEE Security & Privacy, 2021

The Common Vulnerability Scoring System score is the de facto standard to assess risk of software... more The Common Vulnerability Scoring System score is the de facto standard to assess risk of software vulnerabilities, with three temporal components: exploitability, remediation level, and report confidence. We discuss how the latter may be inferred from the first two, pointing practical and conceptual issues in the usage of temporal risk scores.

Journal of Systems and Software, 2020

Abstract Microservices have emerged as an architectural style for developing distributed applicat... more Abstract Microservices have emerged as an architectural style for developing distributed applications. Assessing the performance of architecture deployment configurations — e.g., with respect to deployment alternatives — is challenging and must be aligned with the system usage in the production environment. In this paper, we introduce an approach for using operational profiles to generate load tests to automatically assess scalability pass/fail criteria of microservice configuration alternatives. The approach provides a Domain-based metric for each alternative that can, for instance, be applied to make informed decisions about the selection of alternatives and to conduct production monitoring regarding performance-related system properties, e.g., anomaly detection. We have evaluated our approach using extensive experiments in a large bare metal host environment and a virtualized environment. First, the data presented in this paper supports the need to carefully evaluate the impact of increasing the level of computing resources on performance. Specifically, for the experiments presented in this paper, we observed that the evaluated Domain-based metric is a non-increasing function of the number of CPU resources for one of the environments under study. In a subsequent series of experiments, we investigate the application of the approach to assess the impact of security attacks on the performance of architecture deployment configurations.

Anais do Brazilian Workshop on Social Network Analysis and Mining (BraSNAM), 2019

O tempo que usuários passam nas redes sociais é um tema em voga. A retenção das redes sociais tem... more O tempo que usuários passam nas redes sociais é um tema em voga. A retenção das redes sociais tem importantes implicações, que extrapolam o campo social atingindo aspectos econômicos, psicológicos e de infraestrutura da rede. Neste artigo, consideramos o problema da determinação da taxa ótima de acesso a uma rede social. Para tal, propomos um modelo analı́tico que permite determinar, em função da taxa com que as fontes geram conteúdos, a chance de um usuário acessar a rede e obter conteúdos novos. Em seguida, usamos os resultados do modelo para alimentar um problema de otimização, no qual assumimos que cada acesso à rede social envolve custos. Usando o modelo e o problema de otimização, provemos insights sobre a taxa ótima de acesso. Nossos resultados são parametrizados usando dados do Facebook, que indicam o poder preditivo da abordagem proposta.

Anais do Brazilian Workshop on Social Network Analysis and Mining (BraSNAM), 2018

Facebook news feed personalization algorithm has a significant impact, on a daily basis, on the l... more Facebook news feed personalization algorithm has a significant impact, on a daily basis, on the lifestyle, mood and opinion of millions of Internet users. Nonetheless, such algorithms lack transparency challenging researchers to improve their fairness and accountability. In this paper, we propose a model to capture the dynamics of contents over a timeline (also known as news feed). The input to our model is a fundamental quantity associated to timelines, which we show that can be easily parameterized using real world data: the arrival rate of posts of a given publisher followed by the user. Using real world Facebook traces from the latest elections in Italy, we validate the accuracy of the proposed model and use the model for conterfactual what-if analysis.

Anais do XXXVII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC 2019), 2019

Ataques DDoS são uma grande ameaça. Para mitigar seu impacto, sua detecção deve, preferencialment... more Ataques DDoS são uma grande ameaça. Para mitigar seu impacto, sua detecção deve, preferencialmente, ocorrer na borda da rede, próximo à origem do ataque. Para tal, o uso de roteadores domésticos é uma opção. Entretanto, esses dispositivos são tipicamente limitados, tornando inadequadas as abordagens baseadas, por exemplo, em DPI. Propomos uma abordagem extremamente leve para detecção de DDoS usando roteadores domésticos que emprega exclusivamente contadores de bytes de interfaces de rede. Para detectar ataques com informações tão limitadas, usamos modelos de Machine Learning treinados com dados de centenas de usuários domésticos, obtidos através de parceria com um ISP, juntamente com ataques gerados em ambiente controlado usando as botnets Mirai e BASHLITE. Nossos classificadores são capazes de detectar diferentes vetores de ataque com F1 scores tipicamente superiores a 0,99.