Large scale distributed systems
- 2 ECTS
Large-scale distributed infrastructures leverage the high performance networks to federate computing, data and scientific resources from multiple institutions interconnected through the Internet. Distributed computing technologies have undergone a very fast evolution these last years and the infrastructure deployed have become a critical tool in many scientific disciplines. This lecture describes the foundation of distributed computing infrastructures. It introduces the main computing models exploited in Grids and Clouds to evolve from cluster computing towards more virtualized resources and across-institutional user communities. The main problems encountered when deploying such very large-scale infrastructures are discussed: users identification and authorization, security of data and computations, heterogeneity of resources, redundancy and fault tolerance, deployment, management, and computation flow control… The most wide-spread technologies and their associated middlewares are reviewed. Several examples illustrate the concepts introduced.