Datasets are appropriate when you are working with disconnected data as is the case in many network applications that retrieve database data, manipulate it then merge it back into the datastore. Datasets were designed to facilitate this pattern. Datasets are considered "too heavy" in situations where you are doing read-only access in particular when you are not caching the data. From what you've described, passing datasets seems like the right thing to do.
Dig Deeper on Web data services and enterprise mashups
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.