"Priming these clusters with data from existing client repositories is difficult due to a number of issues including data size, network constraints, security & lack of domain knowledge." . . . . .