[CentOS] [Q} how can O.S. predicate a disk going to failure??

Tue Aug 4 14:02:56 UTC 2009
David.Mackintosh at xdroop.com <David.Mackintosh at xdroop.com>

On Tue, Aug 04, 2009 at 01:38:27PM +0200, Mogens Kjaer wrote:
> > 3. do I need replace this disk now?
> 
> That would be a good idea, the disk could fail in
> 5 minutes or in 5 month, you can't tell.

Or, indeed, 5 years.  I have a number of "throwaway" workstations at
one customer site -- throwaway in that if the disk or system fails,
we just rebuild it, and away it goes.  Several have been telling me
about SMART warnings for YEARS.  My experience seems to echo the
Google study from a few years back, where SMART wasn't an accurate
predictor of disk failure -- some drives SMART then fail, some SMART
for years, and some just fail.

So the answer is "it depends".  If getting a replacement is likely
to be tricky (ie more than a two or three hour wait), or if the data
being stored is highly valuable, then AT LEAST get a spare on site
and sit it next to or on the system in question.  If the data is
extremely highly valuable, do the swap now.

But if you don't care about the data, and/or can tollerate some
downtime, don't worry about it.

Backups *are* good, right?  :)

-- 
 /\oo/\
/ /()\ \ David Mackintosh | 
         dave at xdroop.com  | http://www.xdroop.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.centos.org/pipermail/centos/attachments/20090804/0df04dca/attachment-0005.sig>