In <aanlktinzuyqrsdjpydg+gvl_svs1ubz795fud7k=j...@mail.gmail.com>, Arthur Bela wrote: >HDD#1 >HDD#2 > >I copy files between HDD#1 and HDD#2. > >When i finish, i need a quick "hasing method" - i just want to check, >that the copy was 100% ok.
Hashing is not 100%. It's probabilistic. If your hash results is N bits, your data longer than N bits, and your hash method is uniform, you'll get roughly (1 - 1/2^N) accuracy. For MD5 and SHA1 N = 128, so you'll get 99.99999999999999999999999999999999999971% accuracy. >md5sum, sha256sum is slow -> are there any "very fast" hash algoritms? Not that are cryptographically secure. CRC32 and whatever the UNIX sum program does can catch errors, but are easily deceived by an attacker. >- just for checking if the copied file is corrupt or not [i just need >to know, if there were even 1 Byte error when copying] > >i can write the script, that checks the files, i just need a hash >algoritm [""software/command""], that is fast enough. The slowest part of any such script is not the hash function, it's all the I/O to the disks. Direct byte-by-byte comparison will do the same amount of I/O (reading every byte of every file), and spend no time in a hash routine. There will be a much larger number of comparisons done, but these are very fast compared to the machinations of even a simple hash router. A recursive diff will probably be faster than whatever script you might be writing. -- Boyd Stephen Smith Jr. ,= ,-_-. =. b...@iguanasuicide.net ((_/)o o(\_)) ICQ: 514984 YM/AIM: DaTwinkDaddy `-'(. .)`-' http://iguanasuicide.net/ \_/
signature.asc
Description: This is a digitally signed message part.
