hi all
sequential file:
it act as a source & permanent storage for target.it 's extend is .txt.
dataset:
it act as a temporary storage stage mainly it used before the target stage
while using this stage the ip datas r partited &convert into internal dataset format.then it is easy to load the data in target stage .
copy:
it act as a placeholder.it has a single ip &many o/p .
if u want 2 add a new stage in ur job at that time it is very easy otherwise u have to modify that whole job.