Smartd error

[mike cloaked]

On Mon, Sep 13, 2010 at 4:00 PM, Timothy Murphy <gayleard at eircom.net> wrote:
> Suvayu Ali wrote:
>
>> On Monday 13 September 2010 05:38 AM, Timothy Murphy wrote:
>>> Yesterday I got an internal email from smartd telling me
>>> of an error on one disk (on my server):
>>> ------------------------------
>>> Device: /dev/sdb, 1 Currently unreadable (pending) sectors
>>> ------------------------------
>
snip
> I've started that, thanks.
> I'll see how it goes.
>
> I only have 1 bad sector,
> where I see that you had over 100.
> Still, I'm going to back up my disk onto a USB drive.

I would certainly make sure you have backups as soon as you possibly
can - I have had this in the past and although the disk may run for
many weeks throwing periodic smartd mails or popups at you, it could
fail in a bad way at some point which may be sooner rather than later.

In my case I took a backup as soon as the errors began, and no more!
If the data gets corrupted and then you back up corrupted data it
could lead to problems (understatement).

Again what I did was get a new drive and then did a secure erase on
the drive that was failing (it still was working enough to be able to
do it!) - and then replaced the drive, and partitioned the new drive
from scratch. Then installed a new system on it, before retrieving the
backup data and getting the system operational again.  Total time from
first smartd impending failure messages to new system was a week -
however YMMV.

Good luck.

Mike

-- 
mike c