Abstract
Declustering is a well known technique to achieve high performance for queries on parallel databases. In this paper, we propose new General Disk Modulo (GDM) based declustering algorithms, GDM_Cartesian and GDM_Circle, for distributing uniformly distributed multidimensional datasets to parallel disks, for datasets of any dimension. We compare the performance of the new approaches with several existing declustering algorithms, using variable numbers of disks, and with variable shapes and dimensions of the datasets. Our results show that the new approaches significantly outperform the others for almost all configurations tested.
| Original language | English |
|---|---|
| Pages | 119-127 |
| Number of pages | 9 |
| State | Published - 1999 |
| Externally published | Yes |
| Event | Proceedings of the 1999 International Database Engineering and Application Symposium, IDEAS'99 - Montreal, Que, Can Duration: 02 08 1999 → 04 08 1999 |
Conference
| Conference | Proceedings of the 1999 International Database Engineering and Application Symposium, IDEAS'99 |
|---|---|
| City | Montreal, Que, Can |
| Period | 02/08/99 → 04/08/99 |