|
Re: cyclic redundancy check
Ah, CRC
Cyclic Redundancy Check is one of the two most common forms of hashing, along with MD5. (Message Digest 5). Hashing is essentially a very lossy, very high compression method. By hashing data, a small string of characters called a checksum is produced (that is often represented in the form of hexadecimal values; 0-9 and A-F) that is basically a signature of that data. CRC gets it name from the method of generation: it is a cyclic algorithm to remove redundancy and generate a checksum.
The huge advantage of hashing is that you can transmit or store large volumes of data along with a checksum, so that when the data is next accessed, the same method of hashing can be performed on the data volume and the checksum compared. If they’re the same, that means the data volume should be identical to when the original hash was generated, if they’re not; either the data volume or the hash string has changed; in ether case the data is corrupt in some way and should be discarded.
CRC error or CRC failure normally means that the data volume doesn't match the checksum. However it also mean that the checksum is corrupt (or in bad software, missing), that the hashing method failed for some reason or that the actual data transfer failed. It is very context specific, but in any case, it always means the system believes the data is corrupt, and usually means it really is.
|