Ray Leventhal wrote: > Hi all, > > I perform a nightly snapshot of /home to a USB attached drive scheduled > via cron. The system is CentOS 5.2 and only gets attached to the > internet periodically for updates, otherwise serves as a samba server to > about 20 Windows clients. > > The rsync command being used is: > rsync -av --delete /home/ /media/bkup320G/ > and has been working well until a few days ago. > <snip> Thanks to all who replied. I ran fsck on the disk yesterday and no errors were reported; had already checked cables, etc. About an hour ago I rebooted the machine and let hal mount the attached drive as opposed to mounting it manually, which I'd been doing via script. Also, there was nothing unusual in the logs. I changed my backup script to properly identify the new mount point (/media/disk vs /media/bkup320G) and ran a test on a few home folders which had errored earlier - interestingly, no errors now. So, while my users are offline, I'm rerunning the whole snapshot. I expect it will be fine, though I'm still not sure what got borked up. When I convince the powers that be to get an internal drive for this purpose so that we can do the rsync across the backplane vs USB 2.0, I'll be doing the dance of joy :) Again, thanks to all -Ray