Legacy Formats and Media: A Field Reference for Records Offices

Every records office with more than twenty years of computing behind it has a drawer, cabinet or box of media that nobody can identify with confidence. Some of it is junk. Some of it is the only surviving copy of records you are legally required to keep. The difference is rarely obvious from the outside, and the wrong first move, putting an unknown tape into whatever drive happens to be handy, can destroy what's on it.

This reference is for the first pass: recognizing what you have, understanding what can go wrong with it, and knowing what it takes to read it. It is not a promise that everything can be recovered. Some media can no longer be read by anyone, some can only be read in part, and sometimes recovery costs more than the data is worth. Learning which case you're in early saves money and grief. For the project side of the work, including planning, chain of custody and reconciliation, see legacy data and media conversion: an overview.

Eras below are approximate. They describe when a format was in common office use, not when it was invented or formally retired.

Physical media

Backup and container formats

A backup tape is two problems stacked together: the physical medium and the software format written onto it. Reading the tape gets you a stream of bytes. Making sense of them requires the format.

Database formats

Old databases are often the index to everything else. Without them, a store of images is close to useless.

Image and report formats

Legacy imaging and document-management systems

Systems such as FileNet, OnBase, Laserfiche, Documentum and Optika generally keep images or document objects separate from the database that indexes them, and many early installations wrote to optical libraries. Details vary widely by version and configuration, so don't trust anyone's generalization, this one included, over the system's own documentation. Before planning an extraction, establish four things: where the objects physically live, which database holds the index, whether annotations and redactions are stored separately, and whether your license includes the vendor's export tools. Our article on file formats for long-term preservation covers where the data should land once it's out.

When you find a box of unlabeled tapes

1. Don't load anything. A sticky or damaged tape can be ruined in a single pass, and can foul a drive's heads for the next tape too.
2. Record the find. Photograph the box where it was found and note the location, the date, who found it and anything written on the outside.
3. Ask about legal holds and open requests. If the contents could be subject to either, preserve the box and tell counsel before doing anything else.
4. Move it somewhere cool, dry and secure. Not a hot car, not a damp basement, not the corner of someone's desk.
5. Identify it physically. Measure and photograph each item front and back, and note form factor, barcodes, write-protect settings and any handwriting. Use the tables above.
6. Follow the paper trail. Old backup logs, purchase orders, IT tickets, service contracts and retirees' memories often explain what was backed up, when, and with what software.
7. Check the retention schedule. If the likely contents are long past retention and no hold applies, authorized disposition may be the right answer. Our guide to writing a retention schedule departments will actually follow explains how to document that decision.
8. If you must read it, image first. Use someone with suitable drives and a non-destructive method, who makes a raw image of each tape before trying to interpret it and records what could not be read.
9. Assume it's sensitive. Keep a chain-of-custody log, and destroy media only once authorized, with a certificate.

For wider background on media risk and obsolescence, the Digital Preservation Coalition at dpconline.org is a good place to start.