In this paper we identify and discuss issues that are relevant to the desig
n and usage of databases handling massive amounts of data in parallel envir
onments. The issues that are tackled include the placement of the data in t
he memory, file systems, concurrent access to data, effects on query proces
sing, and the implications of specific machine architectures. Since not all
parameters are tractable in rigorous analysis, results of performance and
bench-marking studies are highlighted for several systems.