19th
International CODATA Conference
Category: Interoperability
Services-based Data Access and Integration
Middleware for Scientific Database
Deting Yang (ydt@cnic.ac.cn)
and Kai Nan (nankai@cnic.ac.cn)
Computer Network Information Center, The Chinese Academy
of Sciences, China
The importance of data access and integration is growing rapidly in the development
of Scientific Database (SDB), which is a long-term project of Chinese Academy of Sciences (CAS) since 1980s. Fields as diverse as
bioinformatics, geophysics, astronomy, medicine, meteorology and particle physics
are facing an exponentially increasing volume of available data. A prime challenge
for Scientific Database is to enable the effective access, integration, analysis
and presentation from the distributed and heterogeneous scientific data, so
that the scientists can exploit its potential. Data grid is a vehicle for achieving
this. According to our experiences in Scientific Database, this paper firstly
summarizes the characteristics of scientific data resources, and investigates
the key requirements for the middleware to enable uniform access and sharing
for multi-discipline data resources. Furthermore, this paper introduces an architecture
built upon Open Grid Service Architecture (OGSA) to facilitate accessing data
resources and constructing truly distributed, scalable, loosely-coupled application.
Some implementations are given in the last section.