Re: Strange problem with badblocks

Mark.Wright_at_bristol.ac.uk
Date: 08/15/04 

Date: Sun, 15 Aug 2004 14:53:18 GMT

P Gentry <rdgentry1@cablelynx.com> wrote:
: So, just how was badblocks invoked in your script?

/sbin/badblocks $PARTITION

I'm aware that this doesnt specify any blocksize, but from a little
googling around that didnt seem necessary any more...

: e2fsck -ccv _invokes_ badblocks -- with the correct block size (and
: start and end blocks?). badblocks may be run without any clues and
: can guess wrong.

: The + you see in the fdsik output simply means that the partition
: boundary does not coincide with a cylinder end point -- ie., it ends
: before/after the end of the cylinder. But since cylinders are only a
: calculated myth, who cares ;-)

: If you really want to monitor your disks' health, you would be better
: running smartd. If your distro did not include it or it's an older
: version, go here:

Actually I already installed smartmontools, but dont have it running by
default yet. Since it didnt come in the distro I was slightly worried
about running it. I'll start it by default now though...

: If it was me, I would tend to go with the e2fsck report of its
: badblocks run -- if you're paranoid, you can always feed those blocks
: to e2fsck with the -l (ell) option. The likelyhood of you ever
: writing to those blocks is pretty small anyway.

OK, I think you've put my mind at rest here, perhaps it was the lack of
parameters for badblocks that is doing this.
Cheers,
Mark 

--