The Replica Consistency problem in Data Grids

Fast and reliable data access is a crucial aspect in distributed computing and is often achieved using data replication techniques. When replicas are modifiable, changes have to be propagated and applied using complex algorithms. In his Ph.D. thesis, done in collaboration with the Italian National Institute of Nuclear Physics (INFN) and the European Organisation for Nuclear Research (CERN), Dr.Gianni Pucciani studies the replica consistency problem in distributed environments and, in particular, in Grid computing and describes the design and implementation of a service that is able to synchronise both files and heterogeneous (different vendors) databases in a Grid environment. Detailed performance results are presented for a practical use case to show how the service can be used together with Oracle Streams to provide multi-tier replication of conditions databases using Oracle and MySQL databases. This work presents ideas and tools useful for researchers working in Data Management and it can be used in master level courses in Computer Science as an interesting case-study for Grid Computing.