Metadata is the information that describes documents in a document library or data repository. At a minimum, it should contain the document’s unique identifier and name. It is better if the metadata includes configuration information, creation dates, document owners, owning projects or official configuration managers who control it and a good description. Data files are the information being managed. Data files may be PDFs, word documents, executable files for software, code modules and entire database backups.

Metadata and data files go hand in hand. However, there is a difference between the two. And there are several situations in which you may have one without the other.

Why Metadata Records May Lack Documents

A document repository may have a beautiful metadata entry. The metadata includes the full document name, a brief description, the employee who created the document, the owning project, when it was last updated, key words associated with the file, its security level or access controls and location in the database structure. Then the user scrolls down on the screen and sees that there is no associated file. This is equivalent to having an entry in the library’s book listing but find that the book is not on the shelf.

This gap may occur when someone creates a new document but has not yet attached the necessary file. Or a database upload of the metadata occurred but the files didn’t transfer. Another cause is the rushed configuration manager who correctly inputs all of the fields but forgets to attach the new document. In some cases, metadata records are created in preparation for scanned images from archives. The physical task of scanning documents and uploading them takes time and money, and that task may be shelved due to more pressing needs. So the database has a metadata record, but the only drawing remains in the paper archives.

When Documents Cannot Be Found Using the Metadata

A more common problem is minimal metadata. Users may simply input the drawing number but leave off a description except for “part drawing”. Keywords may be copied from a metadata template, leaving the drawing impossible to find by a key word search for the document’s contents like “engine manifold” or “32 bit ASIC”. A rushed drafter may create the next revision of the document but leave the new version’s metadata unchanged from the prior version. Then users have trouble finding the new version, though they would find two objects with the old version’s metadata.

Another problem with metadata comes when there is a mass creation of metadata records. The document metadata could contain minimal information left by a mass record migration. The metadata then ends up containing a generic “part record” or “drawing 12345” in the description field instead of useful search criteria like the parent part or summary of the document’s contents. Information can also be lost during a data migration, where the metadata fields do not match up. The “document summary” field information is truncated when shoehorned into a smaller destination field or lost because there was not a properly mapped destination field in the new database.

Files Without Metadata Records

You can have a file without a metadata entry. This typically happens when the file location linkage is broken or a database error loses the connection between the metadata and the file. In this case, the file often can only be found by a system administrator searching for the file on the server.