Static grouping of small objects to enhance performance of a paged virtual memory
ACM Transactions on Computer Systems (TOCS)
A stochastic approach for clustering in object bases
SIGMOD '91 Proceedings of the 1991 ACM SIGMOD international conference on Management of data
Analysis of the generalized clock buffer replacement scheme for database transaction processing
SIGMETRICS '92/PERFORMANCE '92 Proceedings of the 1992 ACM SIGMETRICS joint international conference on Measurement and modeling of computer systems
The LRU-K page replacement algorithm for database disk buffering
SIGMOD '93 Proceedings of the 1993 ACM SIGMOD international conference on Management of data
Principles of static clustering for object-oriented databases
Principles of static clustering for object-oriented databases
The working set model for program behavior
Communications of the ACM
OCB: A Generic Benchmark to Evaluate the Performances of Object-Oriented Database Systems
EDBT '98 Proceedings of the 6th International Conference on Extending Database Technology: Advances in Database Technology
Partition-Based Clustering in Object Bases: From Theory to Practice
FODO '93 Proceedings of the 4th International Conference on Foundations of Data Organization and Algorithms
VOODB: A Generic Discrete-Event Random Simulation Model To Evaluate the Performances of OODBs
VLDB '99 Proceedings of the 25th International Conference on Very Large Data Bases
DBMSs on a Modern Processor: Where Does Time Go?
VLDB '99 Proceedings of the 25th International Conference on Very Large Data Bases
Opportunistic Prioritised Clustering Framework (OPCF)
Proceedings of the International Symposium on Objects and Databases
A study of replacement algorithms for a virtual-storage computer
IBM Systems Journal
| Hi-index | 0.00 |
Despite over 10 years of research into OODBMS design, performance remains as one of the major problems. I/O reduction has proven to be one of the most effective ways of enhancing performance. Two techniques of improving I/O performance of OODBMS are clustering and buffer replacement. Clustering places objects that are likely to be accessed together onto the same disk page and thus reduces I/O. Buffer replacement involves the selection of a page to be evicted, when the buffer is full. The page evictedid eally shouldb e the page needed furthest in the future. Selection of the correct page for eviction results in a reduction in the total I/O generatedb y the system. These two techniques effect the likelihood of a requested object being memory resident in an interdependent way. This fact has been acknowledged in the existing literature. However despite this acknowledgement no existing work investigates this interdependency. This paper makes the first investigation into this interdependency by exploring the effects of ten different buffer replacement algorithms on the performance of two different clustering algorithms. We developed a new family of clustering algorithms that incorporate cache behavior when performing clustering. We term this new family of clustering algorithms, cache conscious clustering (C3). A particular C3 algorithm (GPC3) was tested against a well known and highly competitive clustering algorithm GGP, andthe results show GPC3 out-performs GGP by upto 40% for popular buffer replacement algorithms such as LRU and CLOCK. These results show for the first time clustering should be approached from a cache conscious point of view.