A checksum is a digit representing the sum of the correct digits in a piece of stored or transmitted digital data, against which later comparisons can be made to detect errors in the data.
The process of digitizing archival materials (video tape, audio tape, photographs, scanned manuscripts, etc.) produces thousands of files of various sizes and formats. It is important that each of these files is preserved perfectly, digit by digit, bit by bit, byte by byte. Any change that takes place, even in a single byte, constitutes possible corruption of the original. And this corruption can and does happen over time. Just as with physical objects, digital objects can deteriorate or “rot”. So, it is very important that we find a way to track any change or corruption that may occur. And if corruption is detected the file or digital data needs to be restored to its original form.
Utilizing checksums makes tracking data changes possible
We have developed a program that calculates a checksum for each of the files stored in the Sacred Archives. These checksums are saved in the Archives Database system. Once a year the program is run on the files, calculating the checksum at that time. If a file has changed, even by one byte, over that year the newly calculated checksum will be different than the original checksum retrieved from the Archives Database for that file.
Since every file is backed up in triplicate on digital tapes, if any file corruption is detected by comparing checksums, a copy can be restored from tape. The checksum for the restored copy can also calculated, and if it is the same as the original checksum saved in the Archives Database, the restored file can then replace the file that changed or became corrupted over the course of the previous year. Having three backups of the each file helps to ensure that there will never be a time that a file is lost due to digital corruption.
The possibility of all four files, the one on disk in the Sacred Archives, and the three backups on tape, getting corrupted simultaneously is extremely unlikely – in fact almost impossible.