On Parallel Retrievals Using Distributed Memory Computers

This paper introduces an eecient indexing organization which provides for parallel retrievals. Given a p-processor massively parallel computer, our organization allows a maximum of p retrievals to be performed concurrently and thus achieves an approximate speed up of p in the best case. The index structure is based on the multi-B-tree index structure. It is compatible with current automatic retrieval and query processing techniques. It is also compatible with a variety of le organizations. In addition to an indexed attribute which we refer to as primary attribute, a secondary attribute can be included in the index at no additional cost. Updates are achieved with a high degree of ee-ciency as well. The parallel organization of a multi-B-tree index improves considerably the performance of the index. The method discussed here can be easily applied to standard B+ tree, B* tree and other B tree variations.