There are a couple of possible reasons. The dataCatalog displays various summary meta-data about each file it knows about. For example, merit files are scanned for the number of events they contain, the run numbers represented, file size, etc. This information comes about in a couple different ways: via metadata stored in the pipeline database which is defined with the job producing the data runs, and by a special script called the "crawler" which operates on its own timescale, trolling for new files in the dataCatalog and trying to interpret their content. So, it may be that the crawler has not yet found the data file in question, in which case, most metadata fields will be empty or "0". Another reason might be that the crawler does not yet know how to extract meaningful data from that particular type of file. At this writing (12 Dec 2007) the pipeline crawler knows about the following file types: {merit, digi, recon, relation, mc, svac, cal, ft1}
If you see something that does not fit the above explanation, please report the details to pipelinelist@glast.stanford.edu.