Intensify the I/O Performance of OODBS by Collaboration between Opportunism and Prioritization
Dheeraj Chooramani1,* and Dr. D.K. Pandey2
1,*Research Scholar, Department of Computer Science, JJTU, Rajasthan, India.
2 Director, Dr. Pandey Professional College, Ghaziabad, UP, India
As we all know that clustering has demonstrated to be one of the most effective performance enhancement techniques for object oriented database systems. The bulk of work is done on static clustering, that is re-clustering the object base when the database is off-line. When 24-hour database access is required this type of re-clustering cannot be used. In these cases clustering is required which can recluster the object base while the database is in operation. We believe that most existing on-line clustering algorithms lack three important properties. These include: the use of opportunism to imposes the smallest I/O footprint for re-organization; the re-use of prior research on static clustering algorithms; and the prioritization of re-clustering so that the worst clustered pages are re-clustered first. In this paper we present a opportunistic priority clustering framework in which any existing off-line clustering algorithm can be made on-line and given the desired properties of opportunism and clustering prioritization. Most importantly it allows the created algorithm to have the properties of I/O opportunism and clustering prioritization which are missing in most existing dynamic clustering algorithms. We have used OPCF to make the static clustering algorithms Graph Partitioning and Probability Ranking Principle into dynamic algorithms. We have used the latest version of VOODB 2007 and OCB 1998 in comparison to other researchers.
Keywords: Object-oriented database system, Static clustering, Dynamic clustering algorithms, Prioritization, Opportunism.
