Publicado em: 19/08/2011

Professora Luciana Arantes, dia 19 de agosto, às 15h15, no Auditório Prof Castilhos

Workshop anual do grupo de pesquisa GPPD do INF

Palestra: A Composition Approach to Mutual Exclusion Algorithms for Grid Applications
Palestrante: Luciana Arantes

Local: Auditório Prof Castilhos
Data: 19/08/2011 (sexta)
Horário: 15h15 – 16h45

Resumo:

A Grid comprises of a large number of nodes grouped into clusters. Nodes  within a cluster are often linked by local networks (LAN) while clusters are  linked by a wide area network (WAN). Therefore, Grids inherently present a  hierarchy of communication delays: the cost of sending a message between nodes of different clusters is much higher than that of sending the same  message between nodes within the same cluster.

Distributed or parallel applications that run on top of a Grid usually   require that their processes get exclusive access to some shared resource  (critical region). A mutual exclusion (MUTEX) algorithm then ensures that  exactly one process can access a critical region at a given time. We are  particularly interested in token-based distributed mutual exclusion   algorithms where a unique token is shared among all nodes (processes), and  its possession gives a process the exclusive right to access the critical  region. Several distributed token-based MUTEX algorithms exist in the literature. However, they do not take into account the above-mentioned  hierarchy of communication latency. Therefore, we have proposed a composition approach which allows the combination of any two distributed token-based MUTEX algorithms: one at intra-cluster level and a second one at  inter-cluster level. By using our composition mechanism different algorithms  can be easily “plugged in” on both levels. Furthermore, performance evaluation tests have shown that the good choice for an inter cluster mutual  exclusion algorithm depends on the application behaviour i.e., the frequency  with which the application processes request the shared resource.

Mini-CV:

Luciana Arantes é professora (Maître de Conférences) na Universidade Pierre  et Marie Curie (Paris 6) desde 2001 e membro do projeto Regal, uma  cooperação entre os laboratórios  LIP6 e INRIA.  Ela é graduada em Ciencia  da Computação pela Unicamp, Mestre em Engenharia Elétrica pela Escola  Politécnica da Universidade de São Paulo e Doutora em Ciência da Computação  pela Universidade de Paris 6 (2000).  Sua área de pesquisa se concentra em  algoritmos e sistemas distribuídos hierárquicos ou tolerantes a falhas para  arquiteturas em grande escala, dinâmicos, Grids e Clouds.