Anirban Mondal
Masaru Kitsuregawa
ABSTRACT
In order to provide fast and timely answers to queries in the context of spatial databases and GIS, we present our solution for effective data migration and tuning strategies in shared-nothing parallel spatial databases. Our purpose is to improve the performance of the indexes. Our approach has the following features. First, our scheme is self-tuning, dynamic as well as query-centric and it can adapt to dynamically changing user access patterns. Second, a global distributed R-tree-based indexing method is employed to facilitate effective data migration. Third, unlike traditional partitioning strategies where each processing element (PE) contains data from a single region of space, we allow each PE to store data from multiple and disjoint regions. This minimizes overlap in regions as well as coverage.We implemented the proposed scheme and conducted an extensive performance study on Fujitsu's AP3000 machine with 32 workstations using real datasets. Our experimental results show that our load-balancing strategy can distribute the load effectively across the PEs in the system, thereby reducing response times of incoming queries.
- T. E. Anderson, D. E. Culler, and D. A. Paterson. A case for now network of workstations. IEEE Micro, 15(1):pages 54--64, 1994. Google Scholar
Digital Library
- A. Guttman. R-trees: A dynamic index structure for spatial searching. Proc. ACM SIGMOD, pages 47--57, 1984. Google ScholarDigital Library
- I. Kamel and C. Faloutsos. Parallel r-trees. Proc. ACM SIGMOD, pages 195--204, 1992. Google ScholarDigital Library
- M.L. Lee, M. Kitsuregawa, B.C. Ooi, K.L. Tan, and A. Mondal. Towards self-tuning data placement in parallel database systems. Proc. ACM SIGMOD, pages 225--236, 2000. Google ScholarDigital Library
- A. Mondal, B.C. Ooi, K.L. Tan, M. Kitsuregawa, and M.L. Lee. Query-centric load-balancing using r-trees in shared-nothing spatial databases. Unpublished manuscript, Available from the AUTHORs, 2001.Google Scholar
- C. Zou and B. Salzberg. On-line reorganization of sparsely-populated b+ trees. Proc. ACM SIGMOD, pages 115--124, 1996. Google ScholarDigital Library
Index Terms
- R-tree-based data migration and self-tuning strategies in shared-nothing spatial databases
Recommendations
Parallel R-Tree Spatial Join for a Shared-Nothing Architecture
DANTE '99: Proceedings of the 1999 International Symposium on Database Applications in Non-Traditional EnvironmentsThe growing importance of spatial data has made it imperative that spatial operations be executed efficiently. The most expensive operation is the join for spatial databases. This paper proposes a Replicated Parallel Packed R-tree and its use in ...
Join-queries between two spatial datasets indexed by a single R*-tree
SOFSEM'11: Proceedings of the 37th international conference on Current trends in theory and practice of computer scienceA spatial join, a common query in Spatial Databases and Geographical Information Systems (GIS), consists in testing every possible pair of data elements belonging to two spatial datasets against a spatial predicate. This predicate might be "intersects", ...
Frequency-adaptive join for shared nothing machines
Progress in computer researchAlthough many skew-handling algorithms have been proposed for simple join operations, they remain generally inefficient in the case of θ-join and in the case of multi-join. A new method for self-balancing equi-join operations on shared-nothing (SN) ...
Comments