A Distributed Query Structure to Explore Random Mappings in Parallel

We explore the possibilities to organize a query data structure in the main memories or hard disks of a cluster computer. The query data structure serves to improve the performance of a parallel algorithm for the computation of the structure of a graph induced by a random function. Tradeoffs between different organizations using main memory or hard disks are developed and quantified with parameters. Thus, for concrete cluster systems with concrete parameter values, the best organization can be selected.

• Type: Conference
• Year: 2006
• Where: PDP
• Authors: Jan Heichler, Jorg Keller
• Keywords: Concrete Cluster Systems, Distributed And Parallel Computing, Hard Disks, PDP 2006, Query Data Structure