Well if the PDF are blank after you extracted with ARSDOC GET... then most probably you have
1) The PDF that was archived was blank, so you need to check who produce the PDF and ask them
2) The PDF that you archive was another format, and then converted to PDF... and the conversion went wrong, and you have now a blank PDF.
Another question do you use PDF Indexer? If yes, then it would be interesting, if you still have it, to look at the original PDF file before using PDF Indexer...
Normally, if you are in that situation, the problem is not OnDemand, but the system that produced this PDF. You need to investigate the whole chain before OnDemand, because once it is in OnDemand, OnDemand will NEVER change your file. Like we say here... garbage IN, garbage OUT... you retrieve what you archive.
(ok, there is an exception if during the retrieval there are conversion method... but that's not the case here, from what I understand)