Dewan Hasanat M. author Columbia University. Computer Science Stolfo Salvatore author Columbia University. Computer Science Hernandez Mauricio author Columbia University. Computer Science Mok Kui W. author Columbia University. Computer Science Columbia University. Computer Science originator contributor text Technical reports New York Department of Computer Science, Columbia University 1994 In this paper, we present new algorithms to balance the computation of parallel hash joins over heterogeneous processors in the presence of data skew and external loads. Heterogeneity in our model consists of disparate computing elements, as well as general purpose computing ensembles that are subject to external loading. Data skew appears as significant nonuniformities in the distribution of attribute values of underlying relations that are involved in a join. We develop cost models and predictive dynamic load balancing protocols to detect imbalance during the computation of a single large join. Our algorithms can account for imbalance due to data skew as well as heterogeneity in the computing environment. Significant performance gains are reported for a wide range of test cases on a prototype implementation of the system. Computer science CUCS-026-94 http://hdl.handle.net/10022/AC:P:12446 English NNC NNC 2012-02-03 10:43:15 -0500 2012-02-03 10:46:31 -0500 6444 eng