On Fri, May 13, 2005 at 11:35:22PM -0400, Roberto C. Sanchez wrote: > Pollywog wrote: > > When I do a md5sum check on large files (500MB or larger), I get this sort > > of > > error: > > > > [EMAIL PROTECTED]:~/RO$ md5sum RO_Beta_v3.2_Full.zip > > error processing RO_Beta_v3.2_Full.zip: failed in buffer_read(fd): mdfile: > > Input/output error > > > > Does anyone know what is going on? > > It looks like the buffer_read() call is failing :-)
yes, I guess it's just what it says: the underlying hardware has a problem reading the data... Can you copy the file (just to check whether some other read access on the file does fail, too)? > Seriously, you are probably running out of memory/swap. > Try inreasing them, or using split(1) on the file and > sum the smaller chunks. I don't think memory matters much here -- due to the nature of the MD5 algorithm (it's 'incrementally' computed), you won't need to hold the whole file in memory -- if that's what was meant. (Actually, I've successfully computed md5sums of files definitely larger than all virtual memory available on the respective machine...) Almut -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
