On Wednesday 27 February 2008, Stroller wrote: > > Of course, this does not detect a succesful, but somehow corrupted, > > copy > > (which should be exceptionally rare, anyway). > > Well perhaps I'm just being paranoid today. > But how do I know that a successful, but somehow corrupted, copy has > not occurred? > > What makes you confident that these are rare? I don't ask this to be > antagonistic, just to increase my own confidence in the `cp` command.
Ah well, I have no statistics here. But I can say that such a thing has never occured to me in the past (or at least if it occured, I did not notice that). Not a definitive proof, I know; rather, just my experience. You are of course free to not trust me and, if you're truly paranoid, you probably should do so :-) > I have to admit that I haven't run this command and I don't have any > idea what its actual resource usage would be. I guess I'd be happy > with a lower-grade of checksumming, if it would reduce the runtime to > acceptable levels. With md5sum one can be - barring certain malicious > external attacks - quite certain that a copied file is identical to > the original. I would be happy with a "the file's there and it looks > ok" level of confidence. Well, md5deep has already been suggested. If you are content with a lower-grade checksumming, you could write your own script that compares file lenghts and calculate checksums only on the first n and last m bytes of each file, for some reasonable values of n and m (bigger is better, as you guess). This is what backuppc (an excellent backup software) does when it has to decide whether a file has changed (and thus has to be backed up) compared with the copy stored in the backup pool. Read this for more info: http://backuppc.sourceforge.net/faq/BackupPC.html#some_design_issues "The hashing function" paragraph. Do note that (of course) that method is not 100% accurate and might report false negatives if the corruption is in the middle of the file and file length did not change. -- gentoo-user@lists.gentoo.org mailing list
