On Jan 19, 2010, at 11:29 PM, "Jon Lewis" <[log in to unmask]> wrote:
> On Mon, 18 Jan 2010, Johnson,Robert E wrote:
>> Not so much a fix as a preventative: in the old days (on other
>> it was often recommended that you do something like
>> dd if=/dev/hda of=/dev/hda bs=128k
>> once a month or so. This would refresh every sector of the drive in
>> attempt to remap bad sectors while they are still soft errors instead
>> of hard errors, and could be run in the background at low priority.
>> Modern filesystems might optimize this out to doing nothing, though.
> I don't see how "modern filesystems" would optimize that out, since
> dd operating on the devices, you're below the filesystem level.
Yeah, good point. I guess I was more tired than I thought.
> Also, that seems like an incredibly dangerous thing to do on a
> device that
> has RW mounted partitions. What if while reading a portion of /dev/
> blocks change (due to filesystem activity) between dd's read and
> Now you've written old data back to the device and likely corrupted
Yeah, I guess I should be awake when I try to elaborate. Obviously
you'd want to do it on an unmounted filesystem.
But anyway the prophylactic use of dd is something I've seen
recommended a few times.
I'm going back to sleep now.