[ILUG] smartd errors

[John Allen]

Gavin McCullagh wrote:
> Hi,
>
> I'm not getting too much luck online looking for this so I wonder does
> anyone here know.
>
> I installed a new home server recently with two identical seagate disks,
> partitions set up with MD RAID1.  I then recently set up smartd to monitor
> the disks and logcheck to monitor the logs.  I'm getting very regular
> (hourly) logs of the form:
>
> Oct 23 21:06:19 robin smartd[6894]: Device: /dev/hda, SMART Usage Attribute: 195 Hardware_ECC_Recovered changed from 63 to 64
> Oct 23 21:06:19 robin smartd[6894]: Device: /dev/hdc, SMART Usage Attribute: 190 Unknown_Attribute changed from 60 to 59
> Oct 23 21:06:19 robin smartd[6894]: Device: /dev/hdc, SMART Usage Attribute: 194 Temperature_Celsius changed from 40 to 41
> Oct 23 21:06:19 robin smartd[6894]: Device: /dev/hdc, SMART Usage Attribute: 195 Hardware_ECC_Recovered changed from 61 to 62
>
>   
Nothing to worry about here, unless the numbers keep going up 
(especially the ECC)
> and I'm not sure whether to worry about them.  For the first time, I've now
> seen this:
>
> Oct 24 02:36:19 robin smartd[6894]: Device: /dev/hda, SMART Prefailure Attribute: 7 Seek_Error_Rate changed from 64 to 65
>
>   
This is definitely a problem, as this is a Prefailure attr, it signifies 
that the drive is going to fail, when though
is another issue. I have seen a message before indicating that smart 
reckoned the drive will fail within 24 hours,
so when you see that, you'd better have a spare handy.
> which is a little worrying.  As it's RAID1 I should be able to withstand a
> disk failure so I'm not panicking but I'd like to know if I can ignore
> any of these (and tell logcheck to do so too) or if they are all telling me
> something important.
>
> Does anyone have much knowledge of this stuff?
>
> Gavin
>
>