Data degradation

Below are a few digital images illustrating degradation. All images consist of 326,272 bits. The original photo is shown on the left. In the next photo on the right, one bit is changed from 0 to 1. In the next two photos, two and three bits are changed. On Linux systems , the binary difference between the files can be detected using the command 'cmp' (i.e. 'cmp -b bitrot-original.jpg bitrot-1bit-changed.jpg').

•  

  0 bits changed

•  

  1 bit changed

•  

  2 bits changed

•  

  3 bits changed

Data degradation in DRAM can happen when the bit's electrical charge in DRAM disappears, possibly by changing program code or data. DRAM can be altered by cosmic rays or other high-energy particles. Such data degradation is known as a ^[1] . ECC memory can mitigate this type of data degradation.

Data degradation resulting from the gradual destruction of storage media over several years. Reasons vary with media:

• Solid state electronics , such as EPROM , flash memory, and other solid state drives , store data using electrical charges that can slowly leak due to imperfect insulation. The chip itself is not damaged, so overwriting the data once a decade or so prevents data corruption. Overwriting requires an intact copy of the master data. When it is needed, it may turn out that this data is lost.
• , such as hard drives , floppy disks , and magnetic tapes , can experience data corruption when the bits lose their magnetic orientation. Periodic updates by overwriting data can mitigate this problem. In warm / humid conditions, these media, especially those that are poorly protected from the surrounding air, are subject to physical information carrier ^[1] .
• , such as CD-R , DVD-R, and BD-R , may experience data corruption due to media failure. This can be mitigated by storing the discs in a dark, cold place with low humidity. Disks of "archival quality" with an extended life are available, but storage is far from eternal.
• , such as punched cards and punched tapes , can literally . Mylar punched tapes are another approach that is not based on electromagnetic stability.

Most disks, and higher-level systems are prone to a small chance of a fatal failure. With the growing capacity of disks, file sizes, and with an increase in the amount of information stored on a disk, the likelihood of data corruption on the disk and other forms of uncorrected and undetected data corruption increases ^[2] .

High-level software systems can be used to mitigate the risk of such failures by increasing redundancy and implementing integrity checking and self-healing algorithms ^[3] . The ZFS file system is designed to account for most of such data corruption ^[4] . The Btrfs file system also includes data protection and recovery mechanisms ^[5] , as well as ReFS ^[6] .