Re: [smartmontools-support]Tracking down errors

[Bruce A. <ba...@gr...>]

> Short and extended off line tests were failing (Read Failure) noting
> two LBA's.  I ran 'badblocks' (in default mode) on each partition to
> find out where the problems were and during the scanning of one
> partition a 'badblocks' found 16 bad blocks.  During the scan, a bunch
> of errors where logged my SMARTD and Redhat.
> 
> from SMARTD :
> 
> smartd[24190]: Device: /dev/hdg, ATA error count increased from 0 to 116
> SMART Prefailure Attribute: 5 Reallocated_Sector_Ct changed from 253 to 239
> 
> and running 'smartctl -a /dev/hdg'
> 
> Error 116 occurred at disk power-on lifetime: 2235 hours (93 days + 3 hours)
>   When the command that caused the error occurred, the device was in an 
> unknown state.
> 
>   After command completion occurred, registers were:
>   ER ST SC SN CL CH DH
>   -- -- -- -- -- -- --
>   40 51 02 33 49 e4 e0  Error: UNC 2 sectors at LBA = 0x00e44933 = 14960947
> 
> 
> And then there were lots of kernel messages like:
> 
> Nov  8 11:29:29 tan kernel: hdg: dma_intr: status=0x51 { DriveReady 
> SeekComplete Error }
> Nov  8 11:29:29 tan kernel: hdg: dma_intr: error=0x40 { 
> UncorrectableError }, LBAsect=283396403, high=16, low=14960947,
> 
> Since running 'badblocks' the disk has logged no errors and things a working 
> fine.

That's good: smartmontools and badblocks have done their job.

> Can anyone shed some light on what I am experiencing?

Your disk had some unreadable sectors.  This happens -- but keep an eye on
it since if this continues to happen something may be wrong with your
disk.

> I am not sure why the SMARTD self tests did not find these errors prior to 
> running 'badblocks'.

It did.  But SMART self-tests only report the first failing block
found.  In your case there were more failing blocks.

> Is my disk worse off now after running 'badblocks'??  Did I create
> problems or just find them?

Even better, you repaired the problems (or more precisely, your disk
repaired them).  The bad sectors on the disk have now been replaced with
undamaged sectors.  But be sure to run self-tests on a regular basis (see
-s Directive in smartd.conf) to watch for additional failing sectors.

Cheers,	
	Bruce