Object Database Scalability for Scientific Workloads

Julian J. Bunn, Koen Holtman and Harvey B. Newman (1999) Object Database Scalability for Scientific Workloads. Technical Report. California Institute of Technology. [CaltechCACR:CACR-1999-177]

Full text available as:

PDF - Requires Adobe Acrobat Reader or other PDF viewer.

Abstract

We describe the PetaByte-scale computing challenges posed by the next generation of particle physics experiments, due to start operation in 2005. The computing models adopted by the experiments call for systems capable of handling sustained data acquisition rates of at least 100 MBytes/second into an Object Database, which will have to handle several PetaBytes of accumulated data per year. The systems will be used to schedule CPU intensive reconstruction and analysis tasks on the highly complex physics Object data which need then be served to clients located at universities and laboratories worldwide. We report on measurements with a prototype system that makes use of a 256 CPU HP Exemplar X Class machine running the Objectivity/DB database. Our results show excellent scalability for up to 240 simultaneous database clients, and aggregate I/O rates exceeding 150 Mbytes/second, indicating the viability of the computing models.

EPrint Type:	Monograph (Technical Report)
Subjects:	All Records
ID Code:	26
Deposited By:	Sarah M. Emery
Deposited On:	18 March 2004
Record Number:	CaltechCACR:CACR-1999-177
Official Persistent URL:	http://resolver.caltech.edu/CaltechCACR:CACR-1999-177
Usage Policy:	You are granted permission for individual, educational, research and non-commercial reproduction, distribution, display and performance of this work in any format.

Archive Staff Only: edit this record