Catastrophic disk failure, where was smartd?

[Les Mikesell]

Roger Heflin wrote:
> 
>>> The big issue is that most of the smart implementations don't scan 
>>> the disk for bad blocks, and in my experience several years ago with 
>>> a 1000+ disks in services was that the #1 failure was bad blocks, and 
>>> smart did little to catch that.    The #2 failure was failure to spin 
>>> up at all, but this seemed to be confined to certain batches.
>>
>> Isn't that what the long surface scan test is supposed to do?
>>
> 
> Probably.   I started using dd test before disks and Linux and other 
> oses supported smart.   It works on any disk (or array) whether smart 
> works or not.

That only catches 'hard' errors.  Modern drives have spare sectors and 
the ability to remap soft errors internally, up to a point, before the 
OS knows anything about them.  If the OS (or dd) sees an error, it means 
you've used up the spares or the internal retries weren't able to fix 
it.  The smart interface is supposed to let you know far along you are 
in using up the internal correction and how often soft errors are hidden 
by the retries.  It seems good in theory, and if it predicts the drive 
is going bad you should probably believe it.  But, I think a lot of 
drives fail faster than the internal corrections can handle so you often 
don't get any warning.

-- 
   Les Mikesell
    lesmikesll at gmail.com