There are a number of different factors that contribute to the limitations in ingestion of large datasets. Is the QDC product the full hadoop version or of the single server variety? If you're wanting to ingest / catalog large datasets, the hadoop version is advisable. That said, for the most part : limitations in large file publishes are constrained by QDC product type and then by infrastructure(disk size, network,etc.). There are no hard caps in the product itself.
You might consider opening a support case, in that they could help you triangulate on the cause of this error faster. Hope that helps.