Re: [smartmontools-support] Problem with selective tests?

[John R. <joh...@an...>]

On 14/04/2007 08:04, Sergey Svishchev wrote:
> On Sun, Mar 25, 2007 at 01:52:43PM +0100, John Robinson wrote:
>> On 17/03/2007 11:24, John Robinson wrote:
>>> I just bought a new drive, a Hitachi Travelstar E5K100 40GB. It's 
>>> making a nasty high-pitched buzzing/scraping sound which lasts 4-5 
>>> seconds every 90 seconds or so, but it's working.
>> [...]
>>
>> My supplier replaced it under RMA, and the replacement does this too.
>>
>> Can anyone explain the following parts of the smartctl output to me? 
>> Firstly this error which happens every time the drive is powered up:
>>
>>   10 59 01 01 00 00 a0  Error: IDNF at LBA = 0x00000001 = 1
>>
>>   Commands leading to the command that caused the error were:
>>   CR FR SC SN CL CH DH DC   Powered_Up_Time  Command/Feature_Name
>>   -- -- -- -- -- -- -- --  ----------------  --------------------
>>   29 00 01 01 00 00 a0 00      00:05:40.400  READ MULTIPLE EXT
> 
> Maybe the disk doesn't like when a 48-bit addressing is used to read
> data below 128GB boundary?..  (READ MULTIPLE EXT uses 48-bit addresses
> to reach sectors after 128GB mark, but this time is was used to read
> sector 1).
> 
> Do you run CentOS on this machine?  This could be a driver bug, either.

Yes, CentOS 4 on this machine. Probably about to upgrade to CentOS 5 
soon, though.

>> and secondly this bit about selective tests:
>>
>> Warning! SMART Selective Self-Test Log Structure error: invalid SMART 
>> checksum.
>> SMART Selective self-test log data structure revision number 1
> 
> This could be a firmware bug -- please post output of 'smartctl -l
> selective -r ioctl,2 /dev/hda'

Sorry, too late now, replaced the drive with a Seagate Momentus after 
getting 3 Hitachi Travelstars in a row which all made the awful noise 
and generated the errors I've reported.

Hitachi's response about the noise was "it's supposed to do that, it's 
our Idle Time Function so that the heads don't always fly over the same 
bit of the disc during active idle", but I got no response when I asked 
why it made a noise for several seconds when it was doing something that 
should have taken no longer than a full stroke. Their response about the 
SMART errors was that if the drive passed their own DFT there couldn't 
be a problem.

Cheers,

John.