VERSANT architecture: supporting high-performance object databases

Vlad S. Wietrzyk, Mehmet A. Orgun

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review

3 Citations (Scopus)


Benchmarks, such as the Cattell/001 benchmark, clearly demonstrate that even the slowest object databases are faster than the fastest relational databases when managing complex data structures. Other independently audited benchmarks, such as the University of Wisconsin's 007 benchmark, demonstrate that VERSANT is the clear, overall leader in ODBMS performance. We propose a clustering schedule that may be convenient for any ODBMS; it combines flexibility, dynamics, and efficiency. This paper describes relative partial validation of our approach using VERSANT. The research reported here takes steps towards supplying insights that should provide guidelines to database practitioners and researchers who are interested in applying high-performance, balanced clientserver object database management systems relying on the data-shipping paradigm.

Original languageEnglish
Title of host publicationProceedings - IDEAS 1998
Subtitle of host publicationInternational Database Engineering and Applications Symposium
Place of PublicationLos Alamitos, California
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages9
ISBN (Print)0818683074, 9780818683077
Publication statusPublished - 1998
Event1998 International Database Engineering and Applications Symposium, IDEAS 1998 - Cardiff, Wales, United Kingdom
Duration: 8 Jul 199810 Jul 1998

Publication series

NameProceedings - IDEAS 1998: International Database Engineering and Applications Symposium


Conference1998 International Database Engineering and Applications Symposium, IDEAS 1998
Country/TerritoryUnited Kingdom
CityCardiff, Wales


  • Object Database System
  • Clustering
  • Performance
  • Buffering


Dive into the research topics of 'VERSANT architecture: supporting high-performance object databases'. Together they form a unique fingerprint.

Cite this