--On 25. November 2004 13:28 -0500 Abid Saigol <ab...@saigol.biz> wrote:
The server reset and boot, apparently due to power fluctuation -- thus the
corruption. I was able to restore just the imapd file (for now) by
extracting it from the tarball. This fixed the IMAP problem.
Its (nearly) impossible for FreeBSD's UFS to corrupt static files
which have have been opened read-only, regardless of any insane
mount-options like async.
Theory: Some power-spikes caused the harddisk to write random crap
on the disk, and very likely also destroyed other content.
That's why you need to overwrite _every_ file on the disk with a
known good copy via upgrade. Just like after a break-in.
I don't want to do a pakage delete and add as that would wipe out my whole
configuration
You did not check the manpages on the commands and arguments listed
in my last reply. Its all documented there.
and I don't have backups right now.
Yeah, real man dont do backups and their disks never die...
I did not do any special mount instructions, opted pretty much for the
default during installation. I changed some of the diskspace allocations,
but not the file system types or anything.
FreeBSD is safe out-of-the box, UFS with Softupdates (30 seconds
delay) and dir-sync.
I only corrupted a partition once with FreeBSD, but that was FAT
accessed via Samba (Disk with MP3-archive moved from Windoze)
Is there a chkdsk type of utility that I can use on UFS to find out
if there is any other corrupted file?
Its called `fsck` and runs after every boot in the background.
But unless you have checksums of every file there is no way to
detect files with modified content.
Roland