The SQL architecture is a distributed relational database. There are several excellent implementations of this architecture, that are designed to support extremely large databases.
They offer a good compromise between the distributedness, and the
cost of query operations.
In particular the database can be replicated in order to improve
scalability and fault tolerance, as long as the number of
queries operations dominates the number of
writes.
The R-GMA architecture is being developed to address
this problem: the scalability of the architecture is further improved
introducing components that combine data from the database and cache
the results (similar to NWS forecasters).
next: A modular architecture for the Grid
previous: Store observations into LDAP directories
up: Seminar May 16, 2003