book_and_mouse

Archiving on unsuitable media is a major risk for successful archiving.

The issues of media failure or obsolescence have attained great prominence over the years. There are many tales of media being unable to be read or the digital information no longer being capable of interpretation. Some of these can be for data which cannot be re-generated, for example data from the NASA Viking Landers in the 1970s or the BBC Domsday Project. The data was rescued only with significant manual intervention but valuable lessons were learned.

Media obsolescence presents the following challenges to archiving policy:

Media failure :
All forms of media are susceptible to minor or catastrophic failure. This of course depends on the type or media. Removable media are particularly at risk – the tapes saved many years ago may have broken or been stretched, or the magnetic signal degraded.

Optical media are also volatile. CD and DVD technology was not designed for long term storage and the surface can be corroded in an unpredictable way. Physical damage is also highly possible. Whilst advice is available on the best approach to handle optical media it is not really suitable for long term archiving.

Hard drives in servers are also liable to failure and archivists constantly worry about “bit rot” and its ramifications. For all of these media types the consequences of a very small failure can be very significant. Where for paper losing one page will only result in local problems, losing just one bit in an encrypted or compressed file can result in the entire file being unreadable. At best the results of minor bit loss are unpredictable and will require specialist intervention.

Lack of hardware to access media :
Removable media such as tapes, DVDs and CDs all require specific hardware to read them, and it is often surprising how quickly this becomes unavailable. There are many tales of archivists having to go to Ebay to find hardware to read media that is less than 15 years old. Part of the problem is the growing divergence within a certain technology family. Tapes of course require the specific hardware used to write them, but in optical media the variation is also growing – compact disc formats now include CD-ROM, Audio CD, Video CD, CD-R (650Mb), CD-R (700Mb), CD-RW (650Mb), and CD-RW (700Mb), and for DVDs the list is longer. It is not guaranteed that the drives of the future will read all of these variations.

Lack of software to interpret the bits on media :
Having found some media that has not failed and read it onto your current computer system the battle is not yet over. The bits stored on the media need to be interpreted to yield the actual information stored by the application that created the file. They may have been compressed or have a particular format that needs the original algorithm to de-convolute. This may be performed by a separate driver or software package to the one that accesses the media.

Recommendation
There are many approaches offered by storage vendors to ensure safe, replicated storage of important information, both locally or remotely on the cloud. It is recommended that you choose one of these, but the central storage or management of this content is critical. Trusting removable media or even single copies of server disks is risking significant failure.

© Copyright 2012 Tessella plc