[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Failing Hard Drive, or False Alarms?



Hi,

On Wed, Sep 10, 2025 at 01:30:47PM -0400, Bruce Halco wrote:
> Device: /dev/sda [SAT], 8 Currently unreadable (pending) sectors and Device:
> /dev/sda [SAT], 30 Offline uncorrectable sectors These seem to come within a
> day or so of a reboot, but it hasn't been long enough to know if that's a
> red herring.

This can never be ignored and is never a false positive. It means that
the drive knows that 30 of its sectors became unreadable. The data
there, if any — it can still be space outside a filesystem — is lost. If
you have no redundancy (RAID, LVM mirror, btrfs, zfs, …) you may have
lost data.

However…

> I ran "smartctl -t offline /dev/sda", and the eventual result of "smartctl
> -a /dev/sda" shows

Are you sure that this is the same drive and they haven't re-ordered
their names or something? Because I agree the results are contradictory.
The first one is saying sda has 30 uncorrectable sectors, but:
> 
>    SMART Attributes Data Structure revision number: 16
>    Vendor Specific SMART Attributes with Thresholds:
>    ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE
>          UPDATED  WHEN_FAILED RAW_VALUE
>      5 Reallocated_Sector_Ct   0x0033   100   100   005    Pre-fail
>      Always       -       0

No sectors reallocated yet.

>    197 Current_Pending_Sector  0x0022   100   100   000    Old_age
>       Always       -       0
>    198 Offline_Uncorrectable   0x0008   100   100   000    Old_age
>       Offline      -       0

But no pending or uncorrectable sectors here. Attribute 198 is what's
being reported in the first message, yet attribute 198 shows as zero
here.

It's normal for drives to reallocate bad sectors on write, in which case
you would expect the pending and uncorrectable counts to go back to zer
while the reallocated count goes up by the same number.

Eventually the drive runs out of spare sectors and start accumulating
permanently unusable areas.

I suppose it is possible that you have a drive that doesn't report
reallocated sectors properly.

> which seems to indicate no problems.

The long self-test shows that the whole drive is currently readable. The
fact that the drive reported uncorrectable sectors is worrying. If you
don't have redundancy you may need to restore some files from backup. If
any more uncorrectable sectors happen then the drive is probably very
close to total failure.

Thanks,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting


Reply to: