There's an interesting report available from the 1st Workshop on Extremely Large Databases, held at the Stanford Linear Accelerator Center on October 25, 2007. One of the things that makes this report valuable is the way it brings together and compares commercial and (academic) scientific projects that are making use of extremely large (petabyte) databases and traditional relational DBMS software. You can find the report, as well as other materials related to the workshop, at

Clifford Lynch Director, CNI.