[kwlug-disc] Hard disk smartctl errors

[unsolicited]

Common wisdom around here seems to say once you start getting errors, 
it's on its way out, and you probably don't want to trust it.

I believe Cedric has experienced numerous occasions of false positives 
from smart. However, I suspect they didn't present themselves with quite 
this level of issue.

I would probably relegate this to being a backup drive (especially if 
you have more than one redundant backup), scratch, or temporary. e.g. 
File transfer from one system to another. Or if I duplicate archive 
stuff off to two disks, this could be one of them. Once written (once), 
you'll probably be OK to read off of it at least once. And if not - you 
have the duplicate. Use even as a mindi/mondo destination.

If you have a failure mid-transfer with such, at least it's a copy of 
something - the original is safe for you to find another disk to copy to 
instead.

You would know better than I, but I assume you wouldn't want this as one 
of your apache server /tmp or cache drives ... failure of such may take 
some time to detect. And no doubt be most irritating and unpredictable 
when it does.

Your time has value. And Murphy says when it does go (if you keep it 
live), it will be at the worst possible time.

Disks just don't seem to last as long as they used to. Pre-SATA days.

Warranty / schmarranty - when she goes it's a PITA, even if you can get 
a replacement after the fact. These days I take warranty length to 
merely mean expected lifespan plus. Less than 5 years and even the 
manufacturer doesn't trust it as much as those they offer longer 
warranties on - I assume.

CDN$0.02


P.S. If you saw 60 minutes last Sunday, you would have seen that the 
U.S. nuke silo's still use 8" floppies. They studied the security 
aspects of their systems, and such systems are considered more secure 
than more modern storage media/methods.

- but they can't find parts to be able to close/lock the vault doors. Go 
figure. (Early 60's era.)


On 14-04-29 09:04 PM, Khalid Baheyeldin wrote:
> To the hardware gurus out there ...
>
> I have a disk that was misbehaving for a while, and after a lot of trying,
> I managed to get the (relatively unimportant) data off of it.
>
> But wondering if it would be usable, or I just should junk it.
>
> Here is what "smartctl --all /dev/sdb" says:
>
> ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED
> WHEN_FAILED RAW_VALUE
>    1 Raw_Read_Error_Rate     0x000f   111   089   006    Pre-fail
> Always       -       35541008
>    3 Spin_Up_Time            0x0003   095   092   000    Pre-fail
> Always       -       0
>    4 Start_Stop_Count        0x0032   100   100   020    Old_age
> Always       -       976
>    5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail
> Always       -       6
>    7 Seek_Error_Rate         0x000f   071   060   030    Pre-fail
> Always       -       12929300578
>    9 Power_On_Hours          0x0032   087   087   000    Old_age
> Always       -       11517
>   10 Spin_Retry_Count        0x0013   100   099   097    Pre-fail
> Always       -       797
>   12 Power_Cycle_Count       0x0032   100   100   020    Old_age
> Always       -       1010
> 184 End-to-End_Error        0x0032   100   100   099    Old_age
> Always       -       0
> 187 Reported_Uncorrect      0x0032   001   001   000    Old_age
> Always       -       992
> 188 Command_Timeout         0x0032   100   096   000    Old_age
> Always       -       12885164241
> 189 High_Fly_Writes         0x003a   001   001   000    Old_age
> Always       -       124
> 190 Airflow_Temperature_Cel 0x0022   066   053   045    Old_age
> Always       -       34 (Min/Max 25/47)
> 194 Temperature_Celsius     0x0022   034   047   000    Old_age
> Always       -       34 (0 16 0 0)
> 195 Hardware_ECC_Recovered  0x001a   043   034   000    Old_age
> Always       -       35541008
> 197 Current_Pending_Sector  0x0012   100   100   000    Old_age
> Always       -       0
> 198 Offline_Uncorrectable   0x0010   100   100   000    Old_age
> Offline      -       0
> 199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age
> Always       -       4
>
>
> There is no warranty left on it.