[plug] DMESG Output

[Bernd Felsche]

"Craig Foster" <craig at fostware.net> writes:

>> [mailto:plug-bounces at plug.org.au] On Behalf Of Daniel Pearson

>> Have been having a look through my dmesg and I get the following:
>> [285227.084587] hda: dma_timer_expiry: dma status =3D=3D 0x21
>> [285237.076070] hda: DMA timeout error [285237.082256] hda:
>> dma timeout error: status=3D0x58 { DriveReady SeekComplete
>> DataRequest } [285237.096202] ide: failed opcode was: unknown
>> [285237.134045] hda: status error: status=3D0x50 { DriveReady
>> SeekComplete } [285237.141990] ide: failed opcode was:
>> unknown [285237.149855] hda: no DRQ after issuing
>> MULTWRITE_EXT [285237.262284] hda: status timeout:
>> status=3D0xd0 { Busy } [285237.270447] ide: failed opcode was:
>> unknown [285237.278920] hda: no DRQ after issuing
>> MULTWRITE_EXT [285237.325854] ide0: reset: success
>> [285257.318815] hda: dma_timer_expiry: dma status =3D=3D 0x21

>> Having a google doesn't really provide a whole heap of info,
>> its a new drive (200gb Seagate). Has anyone come across this
>> before? Running kernel  2.6.15-23-amd64-server - distro is
>> Ubuntu-server

>Three issues I've seen...

>1) IDE cables.
>     Swap and try again
>2) DMA set too high
>     Seagates seem to suffer the most for me... ***Cautiously test***
>the DMA settings, Multiword, and transfer speed settings on hdparm.

2a) Access method detected in BIOS may be inappropriate. I had FOUR
"failed" drives because of that. Seeting the access method to
"Large" on the last spare drive I had, "fixed" the drive.

I'll have to go back and check the other three condemned drives to
confirm that they are really dead. Fortunately, I haven't yet run a
high-speed drill through the casings and platters. [Quicker but less
satisfying than dismantling and beating the platters with a
ball-pein hammer.]

>3) Drive failure
>     smartctl is your friend... It will query the drive to see if the
>drive reports having difficulties... What most hard drive testing utils
>do.
>     Download the bootable CD from seagate to test the drive. It runs
>linux, so it should be pretty on the ball.
-- 
/"\ Bernd Felsche - Innovative Reckoning, Perth, Western Australia
\ /  ASCII ribbon campaign | "Laws do not persuade just because
 X   against HTML mail     |  they threaten."
/ \  and postings          | Lucius Annaeus Seneca, c. 4BC - 65AD.