Can anybody tell me how the NetApps box performs with sequential reads with a database that has grown slowly over time ? I understand that the WAFL fs does not re-write the modified data block to the same place but writes it anywhere ie. to the current position of the heads. Does this mean that the heads get thrashed around when a full sequential read is performed on the data base and consequentially the reads are slow compared to doing random reads? Any information on this would be gratefully received Rgds Riasat