"ad0: TIMEOUT - WRITE_DMA" type errors with 7.0-RC1

Joe Peterson joe at skyrush.com
Fri Jan 25 12:56:24 PST 2008


Chuck Swiger wrote:
> On Jan 25, 2008, at 11:24 AM, Joe Peterson wrote:
>> ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE       
>> UPDATED  WHEN_FAILED RAW_VALUE
>>  1 Raw_Read_Error_Rate     0x000f   114   071   006    Pre-fail   
>> Always       -       82422948
> [ ... ]
>>  7 Seek_Error_Rate         0x000f   084   060   030    Pre-fail   
>> Always       -       286126605
> [ ... ]
>> 195 Hardware_ECC_Recovered  0x001a   063   046   000    Old_age    
>> Always       -       166181300
> 
> These numbers are quite worrysome-- they should be zero or nearly so  
> in a healthy drive.

It seems to depend on the drive manufacturer.  E.g. this is a Seagate.  Every
Seagate I've ever had (or heard about on the web via smartctl dumps) reports
very large numbers for these values.  I've heard it described that Seagate
shows you the raw numbers (and correctable errors do happen all the time in
all drives).

In Western Digital drives (IIRC), the numbers shown are the ones that *should*
be zero, thereby hiding the low-level errors.

Hard to say if my numbers are "too high", but these "corrected" error counts
are always frighteningly high in Seagates.

						-Joe



More information about the freebsd-stable mailing list